identification and characterization of young, nearby, solar-type stars · 2020. 4. 2. ·...
Post on 01-Oct-2020
1 Views
Preview:
TRANSCRIPT
Identification and characterizationof young, nearby, solar-type stars
Item Type text; Dissertation-Reproduction (electronic)
Authors Mamajek, Eric E.
Publisher The University of Arizona.
Rights Copyright © is held by the author. Digital access to this materialis made possible by the University Libraries, University of Arizona.Further transmission, reproduction or presentation (such aspublic display or performance) of protected items is prohibitedexcept with permission of the author.
Download date 07/02/2021 10:20:28
Link to Item http://hdl.handle.net/10150/280625
IDENTIFICATION AND CHARACTERIZATION OF
YOUNG, NEARBY, SOLAR-TYPE STARS
by
Eric Earl Mamajek
A Dissertation Submitted to the Faculty of the
DEPARTMENT OF ASTRONOMY
In Partial Fulfillment of the Requirements For the Degree of
DOCTOR OF PHILOSOPHY
In the Graduate College
THE UNIVERSITY OF ARIZONA
2 0 0 4
UMI Number: 3145095
INFORMATION TO USERS
The quality of this reproduction is dependent upon the quality of the copy
submitted. Broken or indistinct print, colored or poor quality illustrations and
photographs, print bleed-through, substandard margins, and improper
alignment can adversely affect reproduction.
In the unlikely event that the author did not send a complete manuscript
and there are missing pages, these will be noted. Also, if unauthorized
copyright material had to be removed, a note will indicate the deletion.
UMI UMI Microform 3145095
Copyright 2004 by ProQuest Information and Learning Company.
All rights reserved. This microform edition is protected against
unauthorized copying under Title 17, United States Code.
ProQuest Information and Learning Company 300 North Zeeb Road
P.O. Box 1346 Ann Arbor, Ml 48106-1346
The University of Arizona ® Graduate College
2
As members of the Final Examination Committee, we certify that we have read the
dissertation prepared by E r i c Earl Mamaiek
entitled Identification and Characterization of Young . Nearby.
Solar-Type Stars
and recommend that it be accepted as fulfilling the dissertation requirement for the
Degree of Doctor of Philosophy
I - 04 Michael R. Meyer
"Barnes, Liebert"
Bieging
Dennis Zarftsky
date
7 h! date
~? / 1 1 ° Y date
7// At date
7 / ' / 09 date
Final approval and acceptance of this dissertation is contingent upon the candidate's submission of the final copies of the dissertation to the Graduate College.
1 hereby certify that I have read this dissertation prepared under my direction and recommend that it be accepted as fulfilling the dissertation requirement.
Dissertation Director: Michael R./Meyer " (7 date
3
STATEMENT BY AUTHOR
This dissertation has been submitted in partial fulfillment of requirements for an advanced degree at The University of Arizona and is deposited in the University Library to be made available to borrowers under rules of the Library.
Brief quotations from this dissertation are allowable without special permission, provided that accurate acknowledgment of source is made. Requests for permission for extended quotation from or reproduction of this manuscript in whole or in part may be granted by the head of the major department or the Dean of the Graduate College when in his or her judgment the proposed use of the material is in the interests of scholarship. In all other instances, however, permission must be obtained from the author.
4
ACKNOWLEDGMENTS
First\ I would like to thank my parents, Karen and Ronald Mamajek, for almost three decades of constant encouragement and support. I know they have been waiting a long time for me to be no longer considered a "student". I would like to thank my advisor Michael Meyer for his endless support and patience over the past few years. Mike has a gift for teaching how to be, and think, like a scientist. I suspect many more graduate students will benefit from his gift during his career. I would also like to thank Jim Liebert for our many conversations over the years. Jim is the heart of the astronomy graduate program at Steward, and I thank him on behalf of all of the graduate students who have benefited from his efforts. I would like to thank my previous research advisors for helping me reach where I'm at today: Eric Feigelson, Warrick Lawson, and Larry Ramsey. Among other things, I am particularly indebted to Eric, for taking a chance on me as an undergrad researcher, to Warrick, for teaching me how to be an observer, and to both, for guiding my MSc research projects. I have been inspired, influenced, and helped over the past 2+ decades or by many excellent scientists, educators, and enthusiasts. I would like to thank (in order): AAAP, D. Arnett, J. Bechtold, J. Beiging, Bethel Park school teachers, A. Blaauw, A. Burrows, J. Carpenter, J. Charlton, L. Close, J. Cocke, C. Corbally, J. de Bruijne, P. Deutsch, T. de Zeeuw, D. Eisenstien, X. Fan, T. Fleming, B. Green, R. Griffiths, L. Hillenbrand, D. Hines, P. Hinz, W. Hoffmann, R. Hoogerwerf, B. Januzzi, R. Kennicutt, L. Lebofsky, F. Melia, J. Monroe, Mrs. Nunes, E. Olszewski, P. Pinto, K. Ratnatunga, G. Rieke, C. Sagan, G. Schmidt, D. Schneider, G. Schneider, D. Soderblom, J. Stauffer, M. Stein-metz, P. Strittmatter, S. Strom, M. Sykes, L. Takahashi, G. Wargo, G. Wayne, R. White, S. Wolff, C. Wu, and D. Zaritsky. Among the students and "young" folks at Steward, LPL, and NOAO, I would like to thank everyone who has shared in the many good times over the past five years (a/3 order): J. Bailin, M. Brown, F. Ceisla, C. Cooper, C. Drouet D'Aubigny, K. Eriksen, M. Freed, N. Gorlova, C. Groppi, J. Hinz, V. Ivanov, M. "Fishboy" Kenworthy, S. Kim, K. Knierman, C. Kulesa, J. Lee, A. Leistra, W. Liu, A. Marble, C. Meakin, D. Miller, I. Momcheva, J. Monkiewicz, A. Moro-Martin, J. Moustakas, J. Murphy, D. O'Brien, E. Offer-dahl, B. Oppenheimer, C. Peng, T. Pickering, G. Rudnick, N. Seigler, H.-J. Seo, M. Silverstone, A. Stutz, T. Thompson, M. Turnbull, G. Williams, K. Williams, P. Withers, & P. Young. I warmly thank the excellent people at Steward who have helped keep me on track (a/3 order): M. Cournoyer, C. Diaz-Silva, B. Duffy, J. Facio, J. Flores, J. Fookson, B. Fridena, A. Koski, N. Lauver, K. Morse, S. Thomas. Last, but certainly not least, I thank my love Lissa Miller for putting up with me while 1 finished this thesis. 1 don't know how she did it, but I love her for it.
^It is a ludicrous policy that one can write a thesis of unlimited length, but there is a page limit (1) for acknowledgments - especially when the thesis formatting requires so much empty space!
DEDICATION
This thesis is dedicated to my parents Karen and Ronald Mamajek.
6
TABLE OF CONTENTS
LIST OF FIGURES 9
LIST OF TABLES 10
ABSTRACT 12
CHAPTER 1 INTRODUCTION 14 1.1 From Molecular Clouds to T Tauri Stars 16 1.2 "Weak-lined" T Tauri Stars = Post-T Tauri Stars? 19 1.3 Finding Samples of Post-T Tauri Stars 21 1.4 Circumstellar Disk Evolution 22 1.5 Secular Parallaxes to Young Field Stars 24
CHAPTER 2 FINDING POST-T TAURI STARS IN THE NEAREST OB ASSOCIATION 27 2.1 Introduction 27 2.2 Selection of Candidate Pre-MS Stars 29
2.2.1 The Hipparcos Sample 29 2.2.2 The RASS-ACT/TRC Sample 31
2.3 Observations 39 2.4 Analysis of Spectra 41
2.4.1 Spectral Types and Luminosity Classification ......... 41 2.4.2 Additional Spectroscopic Diagnostics 47
2.5 Defining the PTTS Sample 50 2.5.1 Membership Status 50 2.5.2 Sample Contamination 51 2.5.3 Sample Completeness 63
2.6 The H-R Diagram 64 2.6.1 Photometry 64 2.6.2 Temperature Scale 65 2.6.3 Secular Parallaxes 65 2.6.4 Independent Confirmation of Vdisp — 1 km s '^for Low-Mass
Members 66 2.6.5 Luminosities 73 2.6.6 Evolutionary Tracks 73
2.7 Results 75 2.7.1 Pre-MS Ages and Age Spread 76 2.7.2 Turn-off Ages 79 2.7.3 The Census of Accretion Disks 83 2.7.4 Is There a Sco-Cen OB Subgroup in Chamaeleon? 85
7
TABLE OF CONTENTS — Continued 2.8 Discussion 86
2.8.1 Is There Evidence for Star-Formation Before the Primary Bursts? 87
2.8.2 On-going Star-Formation? 89 2.8.3 Is LCC Older than UCL? 92 2.8.4 A Star-Formation History of Sco-Cen? 94
2.9 Conclusions 96
CHAPTER 3 CONSTRAINING THE LIFETIME OF CIRCUMSTELLAR DISKS IN THE TERRESTRIAL PLANET ZONE: A MID-IR SURVEY OF THE 30-MYR-OLD TUCANA-HOROLOCIUM ASSOCIATION 99 3.1 Introduction 99 3.2 Observations 101
3.2.1 The Sample 101 3.2.2 Data Acquisition 105 3.2.3 Reduction 107
3.3 Results 108 3.4 Disk Models 116
3.4.1 Optically-Thick Disk With Inner Hole 117 3.4.2 Optically-Thin Disk 118 3.4.3 Single-Temperature Zody Disk 123
3.5 Discussion 125 3.6 Conclusions 126
CHAPTER 4 KINEMATIC DISTANCES TO YOUNG FIELD STARS 130 4.1 Introduction 130
4.1.1 Motivation 130 4.1.2 The Local Association 131
4.2 Astrometry Background Material 134 4.2.1 Rectangular Equatorial Coordinates 134 4.2.2 Heliocentric Galactic Coordinates 135 4.2.3 The Normal Triad 137 4.2.4 Convergent Points and Velocity Vectors 138 4.2.5 Testing for Group Membership 140 4.2.6 Calculating a Cluster Parallax 142
4.3 Generalization of the Secular Parallax Technique 143 4.3.1 Velocity Fields 144 4.3.2 The Perpendicular Velocity Dispersion .......... 146 4.3.3 Revised Membership Probability 147 4.3.4 Iterative Algorithm 149
4.4 Sample Selection 150
8
TABLE OF CONTENTS — Continued 4.5 Results 155 4.6 Conclusions & Future Work 161
CHAPTER 5 CONCLUSIONS 163
APPENDIX A THE PRE-MAIN SEQUENCE I;,// SCALE 166
APPENDIX B STANDARDS WITH QUESTIONABLE LUMINOSITY CLASS . . .168
APPENDIX C POLYNOMIAL FITS 170
APPENDIX D MIRAC PHOTOMETRIC CALIBRATION 173
APPENDIX E COMMENTS ON INDIVIDUAL SOURCES 179 E.l [PZ99] J161411.0-230536 179 E.2 HD143006 180
REFERENCES 181
9
LIST OF FIGURES
2.1 "MI6" T, . JF Index vs. Sr IIA4077 Surface Gravity Index for Program
Stars and Standard Stars 43
2.2 Equivalent Width of Ha for Program Stars and Standard Stars ... 49
2.3 Equivalent Width of Li IA6707 for Program Stars 51
2.4 Secular Parallaxes vs. Hipparcos Parallaxes for Association Members 68
2.5 Proper Motion Vector Diagram of ROSAT-Tycho Stars in LCC and
UCL Regions 70
2.6 Cumulative Probability Plots of /i^ for Candidate ROSAT-Tycho
Members of LCC and UCL 71
2.7 Histogram of Isochronal Ages for Mean HR Diagram Point 76
2.8 H-R Diagram for LCC and UCL Pre-MS Stars 77
2.9 The Effects of Magnitude Bias on Subgroup Age Estimates ..... 79
2.10 Histogram of Isochronal Ages for LCC and UCL Pre-MS Stars ... 80
2.11 H-R Diagram of B-type Members of LCC and UCL 83
2.12 Map of Pre-MS and B-type Members of LCC and UCL . 93
3.1 (J — KS) vs. {KG — N) color-color diagram for Tuc-Hor members
and other stars observed in this study. 114
3.2 Optically-thin and thick dust models fit to the MIRAC and IRAS
photometry for a typical Tuc-Hor star (HIP 1481) 130
3.3 A'-band excess versus age for stellar samples of varying ages. . . . 131
4.1 Relation between {x , y , z) and (p , q , f ) axes 140
4.2 Teff vs. EW(Li) distribution for Pleiades members 154
4.3 HRD for PEPS Local Association Members 162
D.l Transmission profiles for the MIRAC L, N , 11.6, and Q., filters. . . . 178
10
LIST OF TABLES
2.1 Properties of de Zeeuw et al. (1999) Sco-Cen G-K Candidate Members 31 2.1 Properties of de Zeeuw et al. (1999) Sco-Cen G-K Candidate Members 33 2.2 Properties of RASS-ACT/TRC Candidates 35 2.2 Properties of RASS-ACT /TRC Candidates 36 2.2 Properties of RASS-ACT/TRC Candidates 37 2.2 Properties of RASS-ACT/TRC Candidates 38 2.2 Properties of RASS-ACT/TRC Candidates 39 2.2 Properties of RASS-ACT/TRC Candidates 41 2.3 Spectral Standard Stars . 44 2.4 Stellar Classification Scheme 54 2.5 LCC & UCL Pre-Main Sequence Members from de Zeeuw et al.
(1999) List 55 2.5 LCC & UCL Pre-Main Sequence Members from de Zeeuw et al.
(1999) List . 56 2.6 Pre-Main Sequence LCC & UCL RASS-ACT /TRC Members .... 57 2 . 6 P r e - M a i n S e q u e n c e L C C & U C L R A S S - A C T / T R C M e m b e r s . . . . 5 8 2 . 6 P r e - M a i n S e q u e n c e L C C & U C L R A S S - A C T / T R C M e m b e r s . . . . 5 9 2 . 6 P r e - M a i n S e q u e n c e L C C & U C L R A S S - A C T / T R C M e m b e r s . . . . 6 0 2.6 Pre-Main Sequence LCC & UCL RASS-ACT/TRC Members .... 62 2.7 RASS-ACT/TRC & Hipparcos Stars Rejected as Sco-Cen Members . 63 2.8 Estimates of a-(/i^) from Monte Carlo Simulations and Observa
tions 73 2.9 Age Estimates of LCC & UCL 81
3.1 MIRAC Observations of Tuc-Hor Members 104 3.2 MIRAC Observations of Other Stars 105 3.3 Age Estimates for Tucana-Horologium Association 106 3.4 A'-band Photometry of Tuc-Hor Members 112 3.5 Measured Photometry for Other Stars 113 3.5 Measured Photometry for Other Stars 115 3.6 Model Parameters for Tuc-Hor Members 121 3.7 Effects of Changing Adopted Values on Model Output ....... 126
4.1 Parameters for the "Local Association" 135 4.2 Sample of FEPS Targets Younger than the Pleiades (<125 Myr-old) 156 4.2 Sample of FEPS Targets Younger than the Pleiades (<125 Myr-old) 158 4.3 Kinematically-Selected Local Association Members . 161
B.l Revised Luminosity Classes of Standard Stars 172
11
D.l Zero-Magnitude Attributes of MIRAC Photometric Bands 179 D.2 Predicted MIRAC Standard Star Fluxes on CWW system 180 D.2 Predicted MIRAC Standard Star Fluxes on CWW system 181
12
ABSTRACT
Post-T Tauri stars (PTTSs) are low-mass, pre-MS stars which have ceased ac
creting, and are not necessarily near star-forming molecular clouds. Historically,
they have been difficult to identify due to their benign spectroscopic signatures.
With recent all-sky X-ray surveys and proper motion catalogs, it is now possi
ble to find PTTSs in large numbers. The nearest PTTSs will be important tar
gets for future imaging surveys characterizing dust disks and planetary systems
around young solar analogs. The goal of this work is to systematically identify
samples of PTTSs, investigate the evolution of circumstellar disks, to infer the
fossil star-formation history of molecular clouds, and to estimate kinematic dis
tances to young stars lacking trigonometric parallaxes. We present the results of
a spectroscopic survey which identified 110 PTTS members of the nearest OB as
sociation (Sco-Cen). We find that 2/3rds of the low-mass star-formation in each
OB subgroup occurred in <5 Myr, and that only ~1% of solar-type stars with
mean age ~13Myr shows signs of accretion from a circumstellar disk. In or
der to assess how long circumstellar material is detectable around PTTSs, we
conducted a 10//m imaging survey of post-T Tauri members of the ~30-Myr-old
Tuc-Hor association. The goal was to find evidence of either remnant accretion
disks or dusty debris disks with orbital radii of ^10 AU. Combined with data
from other surveys, we conclude that mid-lR emission from warm dust grains in
the terrestrial planet zones around young stars become undetectable compared
to the stellar photosphere for nearly all stars by age ~20 Myr. Lastly, we present
a technique for calculating distances isolated young field stars that currently lack
trigonometric parallax measurements. The technique is a generalization of the
classical cluster parallax method, but can handle anisotropic velocity dispersions
13
and non-zero Oort parameters. Distances and isochronal ages are estimated for a
subsample of PTTSs included in the Formation and Evolution of Planetary Systems
(FEPS) Spitzer Space Telescope (SST) Legacy Science program. The techniques
developed in this thesis will allow one to efficiently conduct a systematic survey
to identify the nearest, youngest stars to the Sun using existing databases.
14
CHAPTER 1
INTRODUCTION
Understanding the origins of stars and planetary systems is one of the most im
portant goals of modem astronomy. We would like to know: how do molecular
clouds form stars and circumstellar disks, and how do circumstellar disks evolve
into planets? The specific goals of this thesis are to (1) investigate the fossil record
of star-formation in the nearest OB association, and (2) study the evolution of
circumstellar environments within 10 AU of young solar-type stars. In order to
achieve these goals, we need to develop and employ techniques for efficiently
selecting young solar-type stars from among field stars. By "young solar-type
stars", we refer to adolescent stars similar in mass to our Sun (FGK stars, ~0.5-
1.5 MQ) with ages between a few Myr (actively-accreting Classical T Tauri stars)
and ~100 Myr (zero-age main sequence, or "ZAMS" stars). We will often refer
to these objects, at least those that lack accretion disks, as "post-T Tauri stars"
(PTTSs) throughout this work.
Identifying stellar adolescents has historically been a thorny problem due to
their lack of obvious spectral signatures in low-resolution optical spectra (e.g.
objective prism Ho: surveys). Two recent breakthroughs have now made it pos
sible to find these objects efficiently: multi-wavelength all-sky surveys (e.g. the
ROSAT X-ray and EUV surveys), and high accuracy proper motion catalogs (e.g.
Tycho-2; Hog et al., 2000a). The rapid rotation of young stars drives a vigorous
stellar dynamo, producing copious, coronal X-ray emission. Young stars are also
bom in groups with small velocity dispersions, and the Galactic potential has not
15
had time to thoroughly scatter their space motion vectors. The convergence of
their proper motions enhances their visibility in proper motion catalogs. We ex
ploit the combination of X-ray brightness and kinematic coherence of the post-T
Tauri stars in order to efficiently identify them for follow-up spectroscopy.
Finding and studying post-T Tauri stars is crucial to elucidating the evolu
tion of circumstellar disks, the evolution of stellar angular momentum, and the
the formation and destruction(?) of planets and planetesimals. Furthermore, the
nearest post-T Tauri stars are arguably the best targets for resolving circumstellar disks,
and identifying young, luminous Jovian planets. While the benefits of studying post-
T Tauri stars are large for stellar astrophysics, their greatest potential value may
well prove to be for planetary science. Understanding the evolution of stars, their
circumstellar disks, and planetary systems, is crucial for placing our solar system,
and Earth, in context.
Post-T Tauri stars also represent the recently-finished end-products of the star-
formation process. In this context, the age spread of post-Tauri star populations
can reveal how long the epoch of star-formation lasted within a given molecular
cloud. There currently exists a debate as to whether molecular clouds form stars
and dissipate over a short timescale (<few Myr) or much longer (>10 Myr), and
whether high-mass and low-mass stars form contemporaneously, or sequentially.
The view that molecular clouds form stars over a wide spectrum of masses within
a short timescale (<few Myr), and then are dispersed by stellar winds and super-
novae, appears to be favored. However, more observational data are needed to
confirm this picture.
In what follows, we review the basic star-formation process, post-T Tauri
stars, circumstellar disks and planet formation, and well as problems inherent
in deriving distances to isolated young stars, in order to set the stage for this
16
thesis.
1.1 From Molecular Clouds to T Tauri Stars
Imagine yourself looking down on our Milky Way galaxy from above. Our Sun
would be an inconspicuous faint point of light situated about 8 kpc from the
Galactic center, and roughly half way between two spiral arms (Perseus and
Sagittarius). You would notice that most luminous, blue, OB-type stars, H II
regions, reflection nebulae, and dark (molecular) clouds, are concentrated along
these spiral arms. The Sun appears to be situated in an inter-arm clump of molec
ular clouds and young stellar complexes. This inter-arm spiral arc is often called
the "Orion Arm" or "Local Arm". While most of these bright, short-lived blue
stars and molecular clouds are concentrated near the dense spiral arms, we are
fortunate to have, up close within the Local Arm, many examples of OB associa
tions and molecular clouds within a few hundred parsecs of the Sun. Ironically
the Sun is situated in a local under-density of the interstellar medium (the "Local
Bubble"). The nearest molecular clouds and OB associations are situated roughly
150 pc distant in Taurus-Auriga in the 2nd Galactic quadrant, and towards the
Sco-Cen complex in the 4th Galactic quadrant (including the Chamaeleon, Lu
pus, Ophiuchus, and Corona Australis molecular cloud complexes).
A wealth of observational evidence demonstrates that stars are formed in the
densest portions of molecular cloud complexes (e.g. Kenyon & Hartmann, 1995;
Evans, 1999; Myers, 1999). In dense molecular clouds cores, self-gravity over
comes thermal, magnetic, and turbulent pressure, and collapses to form a flat
tened, rotating disk of gas and dust: a protostar (Shu, Adams, & Lizano, 1987;
Shu et al., 1999, and references therein). The dust surrounding the protostar ra
diates (and re-radiates) much of the thermal energy from gravitational collapse
17
in the form of infrared and millimeter wavelength radiation (e.g. Lada & Wilk-
ing, 1984; Beichman et al., 1986; Motte, Andre, & Neri, 1998). One classification
system of young stellar objects (YSOs) is based primarily on the shape of the
spectral energy distribution (Lada, 1987; Adams, Lada, & Shu, 1987). An evolu
tionary sequence seems to exist whereby the coolest and most heavily embedded
(extincted) objects are younger, and evolve into objects whose spectral energy
distributions (SEDs) resemble reddened stellar photospheres. The material in the
protostellar disk either viscously accretes onto the protostar, is ejected via inter
action with the protostar's bipolar jets or stellar winds, or is photoevaporated by
UV radiation (Hollenbach, Yorke, & Johnstone, 2000). As the extinction along the
line-of-sight between the nascent star and the observer decreases, the young star
becomes visible at optical wavelengths, and is called a "T Tauri star".
T Tauri stars were first identified as a stellar class by their variability and
strong emission lines (Joy, 1945). The case for their classification as newly-formed,
low-mass, pre-main sequence stars comes from the following: (1) proximity to
dark or bright nebulae (both spatially, and in velocity), (2) spatial coincidence
with unbound, short-lived (<tens Myr), OB associations, (3) luminosities above
the main sequence on a Hertzsprung-Russell diagram, and (4) high Li abun
dance^ . T Tauri stars exhibit near-infrared excess emission, first identified and
attributed to circumstellar dust by Mendoza (1968). The most popular model
for explaining the characteristics of T Tauri stars is that of a pre-main sequence
star which accretes gas along magnetic field lines from a truncated circumstel
lar disk (e.g. Koenigl, 1991). Periodic variability can be caused by the rotational
^The very youngest stars have a Li abundance similar to that observed in meteorites in our own solar system (Strom et al., 1989), suggesting that the Galactic ISM [Li/ H] ratio has changed little over its history. The two prevalent isotopes of Li (®Li and ^Li) are efficiently destroyed at ~2xl0® K in stellar interiors. Pre-MS low-mass stars are mostly convective, and can expose the gas in the outer covective layer to zones with the high temperatures and densities required for Li burning. In this sense, the photospheric abundance of Li can be used as a rough chronometer.
18
modulation of dark starspots (magnetically cooled regions) and bright regions
(heated (~10'' K) boundary regions between the star and magnetospheric accre
tion columns). Non-periodic variability is also seen in many T Tauri stars, and
may be due to variations in accretion rate and extinction. Several hundred T Tauri
stars have been found clustered among nearby dark clouds via their variability
or strong emission lines (e.g. Herbig & Bell, 1988).
Comparing the positions of T Tauri stars in the Hertzspnmg-Russell diagram
(HRD; i.e. effective temperature vs. luminosities) to theoretical evolutionary
tracks, one surmises that the majority are less than a few Myr-old (Cohen &
Kuhi, 1979; Hartmann, 2001). As the dark clouds associated with T Tauri stars
are thought to survive for tens of Myr (e.g. Blitz & Shu, 1980) in order to account
for the present star-formation rate in the Galactic disk compared to the number
of molecular clouds observed, it is reasonable to ask: where are the evolutionary
descendants of the T Tauri stars? If ambipolar diffusion is constraining which
parts of a cloud proceed to collapse, presumably a range of initial conditions will
lead to a wide range of formation times for the stars formed in a cloud complex.
Is such a range of age observed?
Herbig (1978) logically pointed out that T Tauri stars must lose their signa
tures of activity (as manifested through IR excess and emission lines) at some
point, separate themselves from star-forming molecular clouds, and join the field
population. If the majority of T Tauri stars were observed to have isochronal ages
of ;^3 Myr, and the ambipolar diffusion timescale is (coincidently) similar in du
ration to the Kelvin-Helmholtz time for a ~1 M© star to reach the main sequence
(~30 Myr), where are the 90% of pre-MS stars which are older (~3-30 Myr) than
the observed T Tauri star populations? There was little evidence of this "missing"
pre-MS stellar population at the time of Herbig's study. Very few candidates were
19
then known which had characteristics intermediate between that of T Tauri stars
(<few Myr) and zero-age main sequence stars (like those found in the Pleiades
cluster; ages ~100 Myr). So where are Herbig's"post-T Tauri" stars?
1.2 "Weak-lined" T Tauri Stars = Post-T Tauri Stars?
Observations with the Einstein X-ray observatory in the early 1980's yielded a sur
prise. Nearby star-forming clouds had the appearance of "X-ray Christmas trees"
(Montmerle et al., 1983). The T Tauri star populations in molecular clouds were
strong, variable X-ray sources, with luminosities roughly 2-4 orders of magnitude
brighter than that of the Sun (e.g Feigelson & Kriss, 1981). Interspersed among the
X-ray-luminous T Tauri stars was another, separate population of X-ray-luminous,
Li-rich, pre-MS stars (Feigelson & Kriss, 1981; Walter & Kuhi, 1981). These new
objects ("weak-lined T Tauri stars" ; WTTSs) lack evidence of accretion from a
circumstellar disk, i.e. strong emission lines, and strong near-IR and UV excess
emission. While many of the WTTSs were co-spatial with the "Classical" T Tauri
stars (CTTSs) near molecular cloud cores, others were widely distributed tens of
parsecs from active star-forming clouds. The isochronal ages of the WTTSs ini
tially suggested a wide range of ages, between ~l-40 Myr (Walter et al., 1988).
The existence of WTTSs with ages of ~1 Myr and CTTSs with ages of ~5 Myr
was already an indication that low-mass stars have a wide range of accretion
disk lifetimes.
From the number density of WTTSs discovered in the Einstein survey fields
around the Taurus-Auriga molecular clouds, the total population of WTTSs was
extrapolated to outnumber that of CTTSs by 10:1 (Walter et al., 1988), in agree
ment with Herbig's predictions. Many believed that the WTTSs found in the
Einstein surveys were the tip of the iceberg of the long-sought post-T Tauri pop
20
ulation. Yet, large scale surveys using spectroscopic or proper motion selection
(Jones & Herbig, 1979; Herbig, Vrba, & Rydgren, 1986; Hartmann et al., 1991)
failed to identify the population predicted to exist by the Einstein studies. Con
flicting data suggested that the post-T Tauri problem was not yet solved.
In the mid-1990's, the ROSAT All-Sky Survey identified hundreds of new X-
ray sources both inside and outside of the nearest molecular clouds. The stellar
counterparts to these X-ray sources are predominantly Li-rich, late-type stars,
and they were originally thought to be the population postulated to exist by Wal
ter et al. (1988) (e.g. Wichmann et al., 1996; Alcala et al., 1995). However, subse
quent studies using high resolution spectroscopy (Wichmann et al., 2000; Covino
et al., 1997), astrometry (Frink, 1999), as well as arguments regarding the star-
formation history of the local Galactic disk (Briceno et al., 1997), suggested that
these objects were a heterogeneous mix of bona fide T Tauri stars associated with
dark clouds, as well as young field stars. At present, there appears to be little
evidence for the existence of substantial numbers of older pre-MS stars (>3 Myr)
centered on existing molecular clouds.
One unexpected result from the ROSAT All-Sky Survey was the discovery of
isolated ~5-30 Myr-old low-mass stars unassociated with known star-forming clouds.
Several of these groups were found within 100 pc of the Sun: the rj Cha clus
ter (Mamajek, Lawson, & Feigelson, 1999), the TW Hya association (Webb et al.,
1999), the /? Pic moving group (Zuckerman, Song, Bessell, & Webb, 2001), and the
Tucana-Horologium association (Zuckerman & Webb, 2000; Torres et al., 2000). It
appears that the molecular clouds that formed these isolated stellar groups have
been destroyed and star-formation has ceased, explaining why they escaped de
tection until recently.
Observations suggest that the formation of stars in molecular clouds is re
21
markably synchronized, with age spreads of <3 Myr within the stellar popula
tions associated with the nearest T Associations (e.g. Ballesteros-Paredes, Hart-
mann, & Vazquez-Semadeni, 1999). The small number of stars with ages >3 Myr
suggested to exist in the nearest molecular clouds can be largely attributed to un
certainties in HRD positions (Hartmann, 2001), and possibly unassociated young
field stars. The classical theory of star-formation suggests that the cloud cores
producing low-mass stars are magnetically supported (Shu, Adams, & Lizano,
1987), and hence the lifetime of molecular clouds should be limited by the am-
bipolar diffusion timescale (~ 10-20 Myr; Pal la & Galli, 1997). While most of the
gas in dark clouds is molecular, a small fraction of the particules are ions, and the
motions of the ions along magnetic field lines impedes the collapse of the molec
ular component. For typical parameters of molecular clouds, the ambipolar dif
fusion timescale is always larger than the free-fall collapse timescale by roughly
an order of magnitude (McKee et al, 1993). The ambipolar diffusion timescale
is controlled by the density of ions and neutral gas, magnetic field strength,
and the diameter of the cloud. It is unclear why this combination of variables
should conspire to produce all stars in a molecular cloud within a few Myr. It has
been proposed that magnetic field pressure may not be the dominant factor con
trolling the star-formation timescale (and hence the average star-formation rate
of the Galaxy), but perhaps collisions between large parcels of mostly neutral
ISM, resulting in the rapid formation of clouds with and large turbulent pressure
(Hartmann, Ballesteros-Paredes, & Bergin, 2001). In order to place quantitative
constraints on this scenario, we need to sample the entire stellar population as
sociated with a fossil giant molecular cloud - including any older PTTSs if they
exist.
22
1.3 Finding Samples of Post-T Tauri Stars
Although strict observational criteria do not exist for classifying Post-T Tauri stars
(PTTSs), a working definition is a low-mass star (<2 MQ) that is Li-rich compared
to stars on the ZAMS such as the ~125 Myr-old Pleiades, and whose theoretical
HR diagram position is above the main sequence (Herbig, 1978; Jensen, 2001).
Since these criteria would also include CTTSs and WTTSs, one could argue that
PTTSs should have ages >3 Myr, the typical lifetime of accretion disks that are re
sponsible for CTTS activity. The lack of observational evidence for PTTSs within
existing molecular cloud complexes (Hartmann, Ballesteros-Paredes, & Bergin,
2001) suggests that if they exist, they are outside of these clouds interspersed
among the field population. Classifying PTTSs by these criteria has complica
tions: (1) few young field stars not associated with well-studied molecular clouds
currently have accurately measured distances (i.e. known luminosities), (2) un
resolved binarity can make stars with known distances appear more luminous,
and thus younger, (3) there is a dispersion in observed Li abundances among
stars with the same masses and ages in coeval open clusters, and (4) other rare
objects, such as evolved, tidally-locked binaries (RS CVns) can lie above the main
sequence, be Li-rich, and X-ray bright, masquerading as PTTSs. X-ray surveys
{Einstein and ROSAT) have proven to be the most efficient means of first identi
fying pre-MS candidates to date. Ideally, one would like additional criteria for
separating bona fide young stars from other species of X-ray-luminous objects.
In §2 we describe a unique survey undertaken in the Sco-Cen OB association
for identifying the largest and nearest sample of solar-type, post-T Tauri stars
ever assembled. We predicted that low-mass Sco-Cen members should be X-ray
luminous and have proper motions parallel to those of the higher mass members
found by Hipparcos. The near-simultaneous availability of an all-sky catalog of
23
X-ray sources {ROSAT All-Sky Survey) and relatively deep proper motion catalog
(Tycho) made it possible to not only isolate active stars from the field population,
but select only those whose motions were convergent with the B-type subgroup
members. Selection by X-ray emission and astrometry proved very fruitful - over
90% of the candidates selected proved to be Li-rich, pre-MS stars. Our study
was able to place constraints on the star-formation history of the Sco-Cen OB
subgroups, as well as determine the incidence of accretion disks around older
pre-MS stars (~ 10-15 Myr).
1.4 Circumstellar Disk Evolution
One important question that can be addressed with representative samples of
PTTSs is the timescale for disk evolution with a few AU. For example, it is thought
that the bulk of the mass accretion for the solar system's terrestrial planets took
place during the first ~10"-10® yr of its history (Agnor, Canup, & Levison, 1999;
Chambers, 2001). From observations of young stars in nearby molecular clouds,
we now know that circumstellar disks are a nearly ubiquitous by-product of the
star-formation process. Most low-mass stars in the youngest star-formation re
gions (e.g. the ~l-Myr-old Orion Nebula Cluster) have spectroscopic or photo
metric evidence of a circumstellar disk (Hillenbrand et al., 1998). The masses of
circumstellar disks found around some T Tauri stars are similar to that of the min
imum mass solar nebula (Beckwith et al., 1990) - the amount of H, He, and metals
required to form the planets in our solar system. The physical sizes of circumstel
lar disks around T Tauri stars are similar to that of our own solar system (lOs-lOOs
AU; McCaughrean & O'Dell, 1996). Considering their masses, dimensions, and
appearance at the very earliest stages of stellar evolution, these disks are hypoth
esized to be "protoplanetary". Radial velocity surveys of nearby solar-type stars
24
indicate that at least ~5% have at least one Jupiter-mass planet orbiting within
a few AU (Marcy & Butler, 2000). We conclude that the formation of gas giant
planets is one likely outcome of circumstellar disk evolution.
The incidence of inner protoplanetary accretion disks diminishes with age,
being very common at ages <1 Myr, and very rare at >10Myr (e.g. §2). The
fraction of low-mass stars with disks, as manifested by IR excess (in the L-band;
3.5/mi), diminishes with age, with half losing their inner disks (^0.1 AU) within
~3 Myr (e.g. Haisch, Lada, & Lada, 20Gla). Using population statistics of pre-MS
stars in the Taurus molecular clouds, multiple studies have demonstrated that the
transition time for disks inside of ~1 AU to go from optically-thick to optically-
thin is ~10® yr (Skrutskie et al., 1990; Wolk & Walter, 1996).
While some studies argue that accretion terminates by age ~6 Myr (Haisch,
Lada, & Lada, 2001a), others suggest that it may continue at lower accretion rates
around some stars until at least ~10 Myr (§2, Muzerolle et al., 2000; Lawson et al.,
2002; Lawson, Lyo, & Muzerolle , 2004). There are preliminary indications that
disks may persist longer for stars in associations which lack massive OB stars
and for the lowest mass stars (e.g. Lyo et al., 2003; Haisch, Lada, & Lada, 2001b).
The question of whether cooler, remnant disks at larger orbital radii (;^1 AU) ex
ist around stars that have stopped accreting, remains unanswered. How much
dust exists within a few AU of post-T Tauri stars, and how does it compare with
models of terrestrial planet formation?
In §3, we present the results of a mid-IR survey of members of a newly-
discovered post-T Tauri star association (Tucana-Horologium) in order to search
for evidence of circumstellar dust. Although it appears that accretion disks (and
the near-IR excesses that accompany them) disappear by age ~10Myr, our sur
vey, combined with the results of other recent surveys, suggest that detectable
25
10 iim emission from either accretion disks or debris disks, becomes very rare
after ~20 Myr. Our survey finds that typical ~3()-Myr-old have less than a few
thousand times the dust grain surface area that our inner solar system has today.
At these levels, the amount of warm dust is undetectable through photometric
modeling of spectral energy distributions.
1.5 Secular Parallaxes to Young Field Stars
While kinematic groups can provide important stellar samples for studying the
star-formation history of a region and circumstellar disk evolution, they might
represent only a small fraction of the integrated young stellar population in the
solar neighborhood. Post-T Tauri stars in the field (outside of known clusters and
associations) are much harder to find and characterize.
Distances to stars are essential if one wants to derive its luminosity, and poten
tially estimate its age through the use of theoretical evolutionary tracks. Direct
determination of distances to stars through the measurement of trigonometric
parallax are limited to stars either bright enough to have been measured by the
Hipparcos satellite (V > 7-9'"^^ ESA, 1997), or "special" enough to be included
in ground-based parallax measurement programs (usually high proper motion
stars, or stars essential for standard candle calibration, e.g. Cepheids). Most iso
lated, young, Li-rich field stars have either poor parallax determinations from the
Tycho experiment on the Hipparcos satellite, or no published values at all. While
most Li-rich field stars are likely to have absolute magnitudes within a few tenths
of the ZAMS, we would like to determine whether any of these field stars are so
young that they are still demonstrably pre-main sequence. Very nearby pre-MS
stars are highly desirable for studies of the early evolution of stars and circumstel
lar disks (and the search for planets), however they remain a difficult population
26
to identify.
In §4, we attempt to estimate distances to young isolated field stars using a
variation of the cluster parallax technique. The cluster parallax technique is typ
ically used on only the nearest clusters and OB associations (de Bruijne, 1999b;
Madsen, Dravins, & Lindegren, 2002), whose members have large, statistically
s i g n i f i c a n t p r o p e r m o t i o n s , a n d w h o s e i n t e r n a l v e l o c i t y d i s p e r s i o n s ( < 1 k m s " ' )
are much smaller than their heliocentric space motions (~ 10-50 kms~^). To esti
mate distances to young field stars which have accurate proper motions (but lack
trigonometric parallaxes), we exploit the fact that the youngest (^100 Myr-old)
field stars that do have accurate trigonometric parallaxes have similar 3D space
motions with a relatively small velocity dispersion (~5 kms~') compared to their
heliocentric speed (~20 km s~^). These young stars are often called the "Local As
sociation" or "Gould Belt" depending on how restrictive the group is defined in
position, age, and velocity. Although the young field stars do not form a cluster
in the classical sense (i.e. a spatially compact group of co-moving, co-eval stars
formed in the same star-formation event), the coherence of the space motion vec
tors for the Local Association allows us to use similar techniques for selecting
members and estimating distances. We develop an iterative algorithm which can
be used to determine membership probabilities and distances for stars that be
long to groups with previously determined velocity fields (mean space motion,
velocity dispersion, and Oort parameters). We find that many of the Li-rich stars
in the "Formation and Evolution of Planetary Systems" (FEPS) Spitzer Legacy
Science program are likely members of the Local Association. The kinematic dis
tances that we calculate imply luminosities that are either on the ZAMS or pre-
MS. The iterative algorithm developed in §4, along with other spectroscopic and
kinematic selection criteria, will be extremely useful for future surveys intent on
27
systematically identifying the youngest, closest, post-T Tauri and ZAMS stars to
the Sun.
28
CHAPTER 2
FINDING POST-T TAURI STARS IN THE NEAREST OB
ASSOCIATION
The contents of this chapter were previously published in Mamajek, Meyer, & Liebert
(2002) and Mamajek (2003).
2.1 Introduction
Investigations of pre-main sequence evolution have been hampered by a lack
of large samples of well-characterized PTTSs. This deficit has impacted studies
of pre-MS angular momentum evolution (e.g. Rebull et al., 2001; Bouvier et al.,
1997), stellar multiplicity (e.g. Kohler et al., 2000), and circumstellar disk evolu
tion (e.g. Spangler et al., 2001; Haisch, Lada, & Lada, 2001a). The nearest post-T
Tauri stars also provide optimal targets for young exoplanet and brown dwarf
searches (e.g. Lowrance et al., 2000). These objects are much more luminous early
in their evolution, and the closest targets enable characterization of the smallest
orbital radii. With a post-T Tauri population in a nearby OB association, we can
address basic questions such as: How long does star-formation persist in a giant
molecular cloud? What is the duration of the accretion phase for young solar-
type stars?
Identifying a bona fide PTTS sample can be accomplished by searching for
low-mass members of nearby fossil OB associations. The Sco-Cen OB complex
(Sco OB2) is the nearest OB association to the Sun (mean subgroup distances
29
range from 118-145 pc; de Zeeuw et al. (1999), hereafter Z99), and covers roughly
2000 square degrees (~5%) of the sky. The complex is comprised of three kine
matic subgroups (Blaauw, 1946) with nuclear ages ranging from 5 to 15 Myr, a
molecular cloud currently undergoing star-formation (the p Oph complex, WiIk
ing, Lada, & Young , 1989; Blaauw, 1991; de Geus, 1992), and several smaller
cloud complexes in the vicinity (e.g. the Lupus, Corona Australis, Chamaeleon,
Musca, and Coalsack clouds). The three subgroups are Upper Scorpius (US; age
5-6 Myr), Upper Centaurus-Lupus (UCL; age 14-15 Myr), and Lower Centaurus-
Crux (LCC; age 11-12 Myr, de Geus et al., 1989). US has been studied extensively
in recent years (e.g. Preibisch & Zinnecker, 1999, and references therein), however
UCL and LCC have received relatively little attention.
In this work, we investigate the low-mass (<2 M©) membership of the two
oldest Sco-Cen OB subgroups (LCC and UCL) utilizing recently available as-
trometric catalogs {Hipparcos, Astrographic Catalog-Tycho (ACT), Tycho Refer
ence Catalog (TRC), and Tycho-2), the 2 Micron All Sky Survey (2MASS), and the
ROSAT All-Sky Survey (RASS). We conduct a spectroscopic survey of two sam
ples: (1) an X-ray-selected sample of late-type stars from the kinematic candidate
membership lists of Hoogerwerf (2000), and (2) the G-K type Hipparcos members
of the OB subgroups from Z99. In §2.2, we discuss the procedure for selecting
candidate pre-MS stars from both samples, and §2.3 discusses the observations
and assembled database. §2.4 describes the data analysis and characterization of
our stellar sample, and §2.5 discusses the selection of pre-MS stars, sample con
tamination, and completeness. §2.6 describes how we construct an H-R diagram
for the subgroups, and §2.7 presents results regarding the ages of the subgroups,
their age spreads, and the frequency of accretion disks around pre-MS stars. §2.8
30
discusses the star-formation history of LCC and UCL, and §2.9 summarizes the
findings of our survey.
2.2 Selection of Candidate Pre-MS Stars
2.2.1 The Hipparcos Sample
Z99 lists Hipparcos Sco-Cen members that were selected using both de Bruijne's
(1999a) refurbished convergent point method and Hoogerwerf & Aguilar's "spag
hetti" method (1999). Their membership lists contained 31 G-K stars in UCL and
21 G-K stars in LCC (their Table CI). Most of these bright stars have been classi
fied in the Michigan Spectral Survey (e.g. Houk & Cowley 1975), however SIM-
BAD^ reveals that most have been studied no further. We limit the survey to the
30 G-K candidates with Michigan luminosity classes IV or V (see Table 2.1). Stars
with borderline F/G Michigan types were not observed. HIP 63962 and 73777
met the criteria, but were not observed. Z99 estimated the contamination by G-
K-type interlopers of all luminosity classes to be 32% for LCC and 24% for UCL.
Of these 31 stars, 17 also have RASS-BSC X-ray counterparts within 40".
'hltp://simbad.u-s trasbg.fr/Simbad
Table 2.1. Properties of de Zeeuw et al. (1999) Sco-Cen G-K Candidate
Members
(1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11)
Name OB a, 5 (ICRS) fia*, us V {B-V) SpT J H K s notes
HIP Grp. h m s ° " (masyr"' ) (mag) (mag) MSS (mag) (mag) (mag)
57524 LCC 11:47:24.55 -49:53:03.0-33.7±0.8 -10.2±1.0 9.07 0.63±0.02 G3/5VPq2
58996 LCC 12:05:47.48 -51:00:12.1 -37.5±0.8 -11.5±0,8 8.89 0.63±0.02 GlVql
59854 LCC 12:16:27.84 -50:08:35.8 -29.1±1.3 -8.6±1.0 9.34 0.67±0.02 G3Vql
60885 LCC 12:28:40.05 -55:27:19.3-34.9±0.8 -16.0±0.7 8.89 0.64±0.02 Glq4
60913 LCC 12:29:02.25 -64:55:00.6-37.5±1.2 -11.4±0.9 9.04 0.73±0.02 G5Vql
62445 LCC 12:47:51.87 -51:26:38.2 -30.4±1.0 -8.7±1.0 9.52 0.80±0.03 G8/K0Vq3
63797 LCC 13:04:30.96 -65:55:18.5-42.4±1.0 -14.2±0.8 8.48 0.74±0.01 G3Vql
63847 LCC 13:05:05.29 -64:13:55.3-36.5±1.1 -19.5±1.3 9.18 0.73±0.02 G5Vq4
65423 LCC 13:24:35.12 -55:57:24.2-29.0±1.0 -13.0±1.0 9.59 0.66±0.03 (G3w)F7q2
65517 LCC 13:25:47.83 -48:14:57.9-39.0±1.2 -20.3±1.0 9.76 0.60±0.04 K0/2V-i-(G)q3
66001 LCC 13:31:53.61 -51:13:33.1-29.0±1.5 -20.1±1.1 9.84 0.72±0.04 G5/6Vq3
66941 LCC 13:43:08.69 -69:07:39.5-32.7±0.7 -19.8±0.9 7.57 0.74±0.01 G2IV/Vq2
67522 UCL 13:50:06.28 -40:50:08.8-29.3±1.5 -22.9±1.0 9.79 0.67±0.04 GlVq2
68726 UCL 14:04:07.12 -37:15:50.5-16.9±0.8 -16.7±0.7 7.11 0.72±0.02 G3IV/Vq3
71178 UCL 14:33:25.78 -34:32;37.7-28.1±1.6 -28.0±1.4 10.18 0.81±0.06 G8/K0Vq3
72070 UCL 14:44:30.96 -39:59:20.6-20.9±1.5 -24.0±1.4 9.32 0.64±0.03 G3Vq3
74501 UCL 15:13:29.22 -55:43:54.6-16.1±0.8 -24.0±0.8 7.47 0.78±0.01 G2IVql
75924 UCL 15:30:26.29 -32:18:11.6-31.7±2.4 -31.9±2.3 8.80 0.65±0.03 G6Vq2
7.91±0.02 7.59±0.06 7.51±0.02 IRXS J114724.3-495250, TWA 19A
7.67±0.017.37±0.03 7.27±0.01 var(0,06), IIOCS J120547.8-510007
8.04±0.01 7.69rb0.02 7.61±0.03 IRXS J121627.9-500829
7.68±0.02 7.40±0.05 7.28±0.03 var(0.02), IRXS J122840.3-552707
7.60±0.01 7.26±0.017.14±0.01...
7.79±0.01 7.38±0.03 7.25±0.01 var(O.lO), IRXS J124751.7-512638
7.02±0.04 6.79±0.026.71±0.01...
7.78±0.01 7.44±0.03 7.36±0.03 var(0.03)
8.45±0.018.16±0.03 S.lOiO.Ol IRXS J132435.3-555719
8.50±0.02 8.18±0.03 8.08±0.03 var(0.05, P=1.09d),V966 Cen,lRXS J132548.2-481451
8.43±0.018.00±0,03 7.83±0.01 IRXS J133152.6-511335
6.21±0.015.85±0.03 5.77±0.01 var(0.08),CCDM 1343-69084RXS J134306.8-690754
8.58±0.018.30±0.04 8.16±0.02 IRXS J135005.7-405001
5.29±0.00 5.40±0.04 4.92±0.05 CCDM14041-3716
8.52±0.018.06±0.027.94±0.02 var(0.15), V1009 Cen
8.19±0.01 7.93±0.02 7.81±0.01...
5.86±0.015.48±0.025.20±0.01...
7.37±0.01 7.05±0.03 6.92±0.02 IRXS J153026.1-321815
32
2.2.2 The RASS-ACT/TRC Sample
To identify low-mass members of an OB association, one can search for stars
whose proper motions are similar to those of high-mass members. The high-mass
membership and moving group solution for each OB subgroup were determined
by Z99 and de Bruijne (1999b). Thousands of faint stars in the ACT and TRC as-
trometric catalogs^ were identified by Hoogerwerf (2000) as candidate low-mass
LCC and UCL members. A high degree of contamination from interlopers is
expected due to the similarity of the space motions of the subgroups to that of
the Local Standard of Rest, compounded by the low galactic latitude of the sub
groups. The selection of ACT/TRC candidate members is described in detail in
§4 of Hoogerwerf (2000).
The Hoogerwerf ACT /TRC membership lists for LCC and UCL were slightly
modified, and filtered, to produce the final target list. First, we requested from
R. Hoogerwerf (personal communication) candidate membership lists with dif
ferent color-magnitude constraints from that described in Hoogerwerf (2000).
The new color-magnitude selection box is essentially a polygon defined by the
Schmidt-Kaler (1982) empirical zero-age main sequence {{B — V) vs. My) at the
mean distance for each subgroup (Z99), where we take all stars A My = 3 mag
above and AMy = 1 mag below the ZAMS line. Hoogerwerf originally selected
only those stars within A My = 1.5 mag above the ZAMS, however this could in
advertently omit younger members or binaries. The selection box contained 1353
ACT and TRC stars in LCC, and 1874 stars in UCL. In order to target low-mass
solar-type stars with G-K spectral types, we retained only those stars with John-
^The ACT (Urban et al., 1998) and TRC catalogs (H0g et al., 1998) were used for target selection for this project in 1999/2000, however we use the photometry and astrometry from the Tycho-2 catalog (available in 2000; H0g et al., 2000a,b) in the data analysis. The Tycho-2 catalog was a joint USNO/Copenhagen project, and its data supercede the contents of the ACT and TRC catalogs.
Table 2.1—Continued
(1) (2) (3)
Name OB a, 5 (ICRS)
HIP Grp. h m s ° '
(4)
Ma* / (masyr"^)
(5) (6)
V (jB - V)
(mag) (mag)
(7)
SpT
MSS
(8)
J
(mag)
(9)
H (mag)
(10)
Ks (mag)
(11)
notes
76472 UCL 15:37:04.66 -40:09:22.1 -20.2±1.7 -27.6±1.4
77015 UCL 15:43:29.86 -38:57:38.6-20.5d=1.6 -31.7±1.2
77081 UCL 15:44:21.05 -33:18:55.0 -19.1±1.8 -29.5±1.4
77135 UCL 15:44:57.69 -34:11:53.7-20.6±3.3 -25.9±2.5
77144 UCL 15:45:01.83-40:50:31.0-19.4±1.3 -31.1itl.4
77524 UCL 15:49:44.98 -39:25:09.1 -24.5±2.0 -25.2±1.8
77656 UCL 15:51:13.73 -42:18:51.3 -18.0±1.2 -30.0±1.0
79610 UCL 16:14:43.02 -38:38:43.5-14.1±3.4 -29.4±3.2
80636 UCL 16:27:52.34 -35:47:00.4-13.1±2.1 -25.5±1.2
81380 UCL 16:37:12.87 -39:00:38.1 -14.4±2.1 -21.5±1.6
81447 UCL 16:38:05.53 -34:01:10.6-11.3±1.5 -23.9±1.0
81775 UCL 16:42:10.36 -31:30:15.0-14.1±1.5 -18.3±1.3
9.39 0.73±0.03 G5Vq2 7.98±0.017.63±0.05 7.52±0.01 IRXS J153706.0-400929
9.66 0.61±0.03 G3Vql 8.61±0,018.34±0.03 8.32±0.02...
9.69 0.75±0.04 G8Vq2 8.27±0.02 7.86±0.03 7.79±0.02...
9.88 0.78±0.02G6/G8IV/Vq3 8.43±0.028.04±0.028.23±0.04CCDM15450-3412,lRXSJ154458.0-341143
9.46 0.57±0.03 GlVql 8.30±0.017.97±0.037.88±0.02var(0.11),lRXSJ154502.0^05043
10.64 1.09±0.12 K0(V)q3 8.81±0.018.27±0.028.13±0.021RXSJ154944.7-392509
9.58 0.74±0.04 G8Vq3 8.15±0.03 7.78±0.06 7.67±0.02 IRXS J155113.5-421858
9.24 0.52±0.03 Gl/G2Vql 8.07±0.017.85±0.03 7.98±0.03 CCDM 16147-3839
9.37 0.68±0.03 G6Vq2 8.04±0.017.71±0.017.62±0.01 IRXS J162752.8-354702
9.82 0.66±0.05 G2/5Vq3 8.45±0.02 8.09±0.04 7.99±0.03...
9.08 0.54±0.05 GllV/Vql 7.91±0.02 7.66±0.03 7.55±0.03...
9.44 0.64±0.04 G5Vq2 8.36±0.02 8.37±0.05 8.08±0.04...
Note. — Columns: (1) Hipparcos ID, (2) OB subgroup region, (3) Hipparcos position (ICRS, epoch 2000.0), (4) proper motion components (masyr"!; where
= HaCOsS), (5) Johnson V magnitude, (6) Johnson (B — V) color, (7) Spectral types from the Michigan Spectral Survey (Vol. 1-3); "ql" indicates the flag
"quality = 1" in the MSS catalog, where q=l,2 stars are judged to have reliable spectral types, (8-10) 2MASS JHKs magivitudes, (11) Notes on variability. X-ray
counterparts, and multiple star system name (CCDM) from Hipparcos. "Var.(N,NN)" indicates that the H6 field in the Hipparcos catalog is identified as being
variable in the broad Hp pass-band. The magrutude scatter N.NN in Hp is listed, and period P if found. Near-IR photometry is from the preliminary 2MASS
database. The name is given of RASS-BSC X-ray sources within 40" of Hipparcos stars; many of the variable stars are also ROSAT sources. Astrometric data and
optical photometry are from from the Hipparcos catalog ESA (1997). pJ
34
son (B — V) > 0.58 mag (the unreddened color of GO dwarfs; Drilling & Landolt,
2000). No red (B — V) limit was imposed. After the color-magnitude selection,
we retained only those stars which were identified as kinematic members in both
the ACT and TRC astrometric catalogs. This final color-selection of the ACT/TRC
lists resulted in 785 UCL candidates and 679 LCC candidates.
In order to further filter the target list, we selected only those ACT /TRC can
didates which had ROSAT All-Sky Survey Bright Source Catalog (RASS-BSC)
X-ray counterparts. Voges et al. (1999) cross-referenced the RASS-BSC with the
Tycho catalog and found that 68% of the optical-X-ray correlations were within
13", and 90% of the correlations were within 25". In plotting a histogram of the
separation distance between RASS-BSC X-ray sources and ACT/TRC stars, we
independently find 40" to be an optimal search radius. No constraints on X-ray
hardness ratio were imposed in the target selection. In order to calculate X-ray
luminosities, we assume the X-ray energy conversion factor for the ROSAT PSPC
detector from Fleming et al. (1995). The linearity of this X-ray efficiency relation
spans the temperature range of stellar coronae from inactive subdwarf stars to
extremely active RS CVns and T Tauri stars. Unsurprisingly, the kinematic selec
tion of ACT /TRC stars also selected many of the same stars as in the Hipparcos
sample (HIP 57524, 59854,62445,65423,66001,66941,67522,75924,76472,77135,
77524, 77656, 80636). These stars are retained in the Hipparcos sample (Table 2.1),
and omitted from the RASS-ACT/TRC list (Table 2.2). The final target list of 96
RASS-ACT/TRC stars (40 LCC, 56 UCL) is given in Table 2.2.
Table 2.2. Properties of RASS-ACT/TRC Candidates
(1) (2)
Name OB
TYC Grp.
9212-2011-1 LCC
8625-388-1 LCC
8222-105-1 LCC
8982-3046-1 LCC
8982-3213-1 LCC
8640-2515-1 LCC
8644-802-1 LCC
8644-340-1 LCC
9231-1566-1 LCC
8636-2515-1 LCC
8242-1324-1 LCC
8637-2610-1 LCC
8645-1339-1 LCC
8641-2187-1 LCC
8983-98-1 LCC
8633-508-1 LCC
8238-1462-1 LCC
8234-2856-1 LCC
8633-28-1 LCC
(3) (4) (5) (6) (7) (8) (9) (10) (11) (12) (13) (14)
a, (5 (ICRS) U S V T { B T - V T ) J H K S IRXS sep. X-ray Hrdns. Table 5
m s o I II (masyr~l) (mag) (mag) (mag) (mag) (mag) Name (") (cts~^) HRl #
LCC 12:11:31.43 -58:16:53.2 -36.0±2.6 -8.7±1.8
10.49±0.04 1.09±0.08 8.49±0.03 8.02±0.05 7.79±0.01 J105751.2-691402 10 1.61e-01 0.31 1
10.02±0.03 1.02±0.06 8.09±0.02 7.64±0.04 7.48±0.02 J113209.3-580319 7 8.25e-02 0.03 2
10.35±0.03 0.75±0.05 9.22±0.02 8.93±0.03 8.86±0.03 J113501.5-485011 24 l.Ole-01 0.49
10.09±0.03 0.75±0.05 8.82±0.01 8.50±0.02 8.40±0.02 J120413.3-641837 16 2.77e-01 0.32
9.49±0.02 0.79±0.03 8.05±0.01 7.71±0.01 7.61±0.01 J120448.2-640942 13 3.47e-01 0.46 3
10.77±0.05 0.81it0.08 9.25±0.03 8.84±0.03 8.77±0.03 J120613.9-570215 3 1.44e-01 0.27 4
10.21±0.03 1.26±0.07 8.31±0.01 7.81±0.01 7.66±0.01 J120941.5-585440 5 2.33e-01 0.16 5
10.29±0.04 1.00±0.07 8.58±0.01 8.09±0.03 7.96±0.02 J121131.9-581651 4 2.41e-01 0.09 6
9.23±0.02 0.79±0.03 7.68±0.01 7.31±0.04 7.19±0.02 J121137.3-711032 5 2.68e-01 0.34 7
10.58±0.04 l.llrfc0.08 8.73±0.02 8.25±0.03 8.13±0.02 J121236.4-552037 11 1.83e-01 -0.02 8
10.38±0.03 1.05±0.06 8.69±0.01 8.26±0.04 8.13±0.02 J121434.2-511004 8 1.77e-01 0.32 9
9.73±0.02 0.92±0.04 8.04±0.02 7.66±0.02 7.52±0.01 J121452.4-554704 0 5.65e-01 0.37 10
10.82i:0.05 0.95±0.09 8.45±0.01 7.84±0.01 7.73±0.01 J121828.6-594307 8 1.78e-01 0.39 11
9.96±0.03 1.02±0.05 7.96±0.01 7.47±0.04 7.31±0.01 J121858.2-573713 5 2.93e-01 0.23 12
10.21±0.04 1.01±0.07 8.03±0.01 7.55±0.01 7.38±0.01 J121919.4-645406 14 2.55e-01 0.26 13
9.41±0.02 0.76±0.03 8.08±0.01 7.74±0.02 7.65±0.02 J122116.7-531747 3 4.34e-01 -0.03 14
10.10±0.03 0.90±0.05 8.49±0.02 8.05±0.02 8.02±0.03 J122155.9-494609 3 1.13e-01 0.59 15
10.59±0.04 0.95±0.06 8.78±0.02 8.27±0.03 8.17±0.01 J122204.0-484118 7 l.lOe-01 0.88 16
9.49±0.02 0.72±0.03 8.18±0.01 7.91±0,04 7.78±0.02 J122233.4-533347 2 1.43e-01 0.00 17 OJ oi
#
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
Table 2.2—Continued
(3) (4) (5) (6) (7) (8) (9) (10) (11)
a, 5 (ICRS) f^oi * M(5 VT {Bt - VT) J H Ks IRXS sep.
h m s ° " (mas yr~ ) (mag) (mag) (mag) (mag) (mag) Name (")
LCC 12:48:07.79-44:39:16.8 -39.7d=1.3-18,0±1.3 9.82±0.03
LCC 13:02:37,54 -54:59:36.8
LCC 13:06:40.12 -51:59:38.6
LCC 13:34:20.26-52:40:36.1 -33.4±1.6-21.1±1.3 9.37±0.02
10.94±0.06 0.94±0.09 9.02±0.02 8.53±0.05 8.37±0,03 J122339.9-561628 4
9.98±0.03 1.10±0.05 8.02±0.01 7.51±0.01 7.35±0.01 J123637.5-634446 10
10.50±0.04 1.07±0.07 8.75±0.04 8.28±0.03 8.16±0.02 J123657.4-541217 13
10.21±0.03 0.88±0.05 8.67±0.01 8.23ib0.04 8.13±0.03 J123938.4-573141 3
10.01±0.03 0.81±0.05 8.40±0.01 8.02±0.02 7.89±0.01 J124118.5-582556 2
10.91±0.05 1.40±0.14 8.54±0.01 8.01±0.01 7.87±0.01 J124432.6-633139 16
10.48±0.04 0.86±0.06 8.70±0.02 8.23±0.02 8.10±0.01 J124506.9-474254 4
9.82±0.03 0.93±0.05 8.12±0.01 7.69±0.05 7.51±0.02 J124807.6-443913 4
10.31±0.03 0.92±0.05 S,88±0.01 8.48±0.02 8.39±0.02 J124847.4-563525 13
10.01±0.03 0.98±0.06 8.20±0.02 7.70±0.06 7.56±0.02 J125824.6-702848 4
11.18±0.06 1.02±0.13 9.44±0.02 8.91±0.02 8.78±0.02 J130153.7-530446 29
10.35±0.03 0.72±0.05 8.86±0.01 8.57±0.03 8.48±0.02 J130237.2-545933 4
10.62±0.04 0.91±0.07 8.90±0.01 8.44±0.04 8.27±0.01 J130638.5-515948 17
10.08±0.02 0.70±0.04 8.76±0.01 8.45±0.02 8.42±0.02 J131327.3-600032 13
10.48±0.04 0.96±0.08 8.68±0.03 8.27±0.03 8.10±0.04 J131424.3-505402 4
10.48±0.05 0.72it0.08 8.86±0.03 8.49±0.05 8.39±0.03 J131754.9-531758 18
10.10±0.03 0.77±0.06 8.97±0.03 8.63±0.05 8.55±0.03 J132204.7-450312 10
10.54±0.04 1.08±0.08 8.28d=0.03 7.65±0.02 7.31±0.01 J132207.2-693812 1
9.37±0.02 0.71±0.03 8.04±0.01 7.69±0.04 7.56±0.03 J133420.0-524032 4
10.17±0.04 0.97±0.06 8.47±0.01 8.03±0.04 7.89±0.02 J133758.0-413448 10
Table 2.2—Continued
(1) (2)
Name OB
TYC Grp.
8667-283-1 LCC
8270-2015-1 UCL
8263-2453-1 UCL
7811-2909-1 UCL
7815-2029-1 UCL
9244-814-1 LCC
8285-847-1 UCL
8282-516-1 UCL
7817-622-1 UCL
7813-224-1 UCL
7814-1450-1 UCL
8683-242-1 UCL
8283-264-1 UCL
8283-2795-1 UCL
7305-380-1 UCL
7828-2913-1 UCL
7310-2431-1 UCL
7310-503-1 UCL
7824-1291-1 UCL
7833-2037-1 UCL
(3) (4) (5) (6) (7) (8) (9) (10)
a, S (ICRS) T I I V T ( B T - V T ) J H IRXS
m s o ' " (masyr"') (mag) (mag) (mag) (mag) (mag) Name
LCC 13:43:28.53-54:36:43.5 -44.0±1.8-20.8±1.8 9.39±0.02 0.77±0.03 8.03±0.01 7.68±0.01 7.63±0.02 J134332.7-543638 36 4.81e-01
8270-2015-1 UCL 13:47:50.55^9:02:05.5 -24.3±2.2-14.7±2.0 10.91±0.07 0.90±0.11 9.30±0.01 8.81±0.05 8.70±0.03 J134748.0-490158
8263-2453-1 UCL 13:52:47.80-46:44:09.2 -22.7±1.3-18.9±1.4 9.69±0.03 0.72±0.04 8.41±0.02 8.07±0.04 7.94±0.01 J135247.0-464412
7811-2909-1 UCL 14:02:20.73-41:44:50.8 -27.6±2.0-19.4±1.9 10.80±0.06 0.90±0.10 8.99±0.02 8.55ib0.04 8.42±0.03 J140220.9^14435
7815-2029-1 UCL 14:09:03.58-44:38:44.4 -20.9±1.1-22.6±1.1 9.46±0.02 0.71±0.04 8.25±0.01 7.98±0.05 7.86±0.02 J140902.6-443838 12 1.73e-01
LCC 14:16:05.67-69:17:36.0 -29.4±2.4-16.0±2.3 10.21±0.03 0.72±0.06 8.74±0.01 8.39±0.02 8.36±0.02 J141605.3-691756 20 6.10e-02
UCL 14:16:57.91-49:56:42.3 -23.2±1.1-22.1±1.1 8.77±0.01 0.71±0.02 7.43±0.02 7.12±0.02 7.03±0.01 J141658.4-495648
UCL 14:27:05.56-47:14:21.8 -25.5±1.7-21.2±1.6 10.68±0.05 0.95±0.08 9.06±0.03 8.65±0.03 8.52±0.03 J142705.3-471420
UCL 14:28:09.30-44:14:17.4 -18.4±2.0-20.6±1.9 9.87±0.03 0.90±0.05 8.31±0.01 7.93±0.03 7.78±0.03 J142809.6-441438 20 1.45e-01
7813-224-1 UCL 14:28:19.38-42:19:34.1 -19.7±1.5-19.8±1.7 10.55±0.05 0.79±0.09 8.92±0.01 8.51±0.03 8.40±0.03 J142817.6-421958
7814-1450-1 UCL 14:37:04.22-41:45:03.0 -21.5±2.0-19.1±1.9 9.77±0.02 0.72±0.04 8.26±0.01 7.90±0.02 7.80±0.03 J143704.6-414504
UCL 14:37:50.23-54:57:41.1 -18.3±2.5-25.8±2.4 10.80±0.06 0.77±0.09 8.94±0.01 8.46±0.02 8.32±0.02 J143750.9-545708
UCL 14:41:35.00-47:00:28.8 -29.6rtl.3-27.4±1.2 10.09±0.03 0.88±0.04 8.43±0.01 7.98±0.03 7.88±0.03 J144135.3-470039 11 6.12e-01
8283-2795-1 UCL 14:47:31.77-48:00:05.7 -29.6±2.3-16.7±2.3 10.79±0.06 0.75±0.09 9.37±0.01 9.04±0.01 8.96±0.03 J144732.2-480019 14 7.37e-02
UCL 14:50:25.81-35:06:48.6 -18.2±2.6-15.8±2.6 10.83±0.07 0.97±0.13 8.75±0.03 8.24±0.02 8.11±0.03 J145025.4-350645
7828-2913-1 UCL 14:52:41.98-41:41:55.2 -22.6±2.2-21.4±2.1 11.02±0.09 1.35±0.22 9.05±0.02 8.40±0.02 8.29±0.03 J145240.7-414206 17 1.30e-01
7310-2431-1 UCL 14:57:19.63-36:12:27.4 -26.6±1.6-22.3±1.7 10.36±0.05 0.97±0.09 8.78±0.03 8.44±0.04 8.30±0.03 J145720.4-361242 17
7310-503-1 UCL 14:58:37.70-35:40:30.4 -21.4±2.2-26.0±2.3 10.88±0.07 1.37±0.20 8.65±0.01 8.08±0.03 7.90±0.01 J145837.6-354036
7824-1291-1 UCL 14:59:22.76-40:13:12.1 -26.0±1.4-26.5±1.6 9.80±0.04 0.86±0.07 8.34±0.01 7.97±0.04 7.82±0.02 J145923.0^01319
10:51.88-43:31:21.0 -20.1±2.1-19.1±2.0 11.23±0.08 0.81±0.13 9.31±0.03 8.88±0.02 8.75±0.03 J150052.5^33107 15 7,57e-02
(11) (12) (13) (14)
Sep. X-ray Hrdns. Table 5
(") (cts~^) HRl #
36 4.81e-01 0.22 37
26 1.50e-01 -0.10 38
8 1.96e-01 0.67 39
15 9.68e-02 0.37 40
12 1.73e-01 0.67 41
20 6.10e-02 0.47 42
7 1.59e-01 0.15
3 1.53e-01 -0.12 43
20 1.45e-01 0.13 44
31 2.23e-01 0.34 45
4 1.68e-01 0.58 46
33 2.20e-01 0.07 47
11 6.12e-01 0.15 48
14 7.37e-02 0.50 49
6 9.03e-02 0.35 50
17 1.30e-01 0.21 51
17 8.67e-02 0.69 52
5 8.30e-02 0.60 53
7 1.87e-01 0.35 54
15 7,57e-02 0.39 55 OJ V]
Table 2.2—Continued
(1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11) (12) (13) (14)
Name OB A , 6 (ICRS) N S V T { B T - V T ) J H K S IRXS sep. X-ray Hidns. Table 5
TYC Grp. h m s ° i " (masyr"^) (mag) (mag) (mag) (mag) (mag) Name (") (cts~^) HRl #
7829-504-1 UCL 15:01:11.56-41:20:40.6 -14.9±1.7-15.1±1.7 10.09±0.03 0.77±0.06 8.78±0.03 8.47±0.05 8.33±0.03 J150112.0-412040 4 1.78e-01 0.41 56
8297-1613-1 UCL 15:01:58.82-47:55:46.4 -22.2±1.7-18.2±1.6 10.22±0.D4 0,71±0.07 8.91±0.01 8.59±0.02 8.51±0.02 J150158.5-475559 13 5.07e-02 0.42 57
7319-749-1 UCL 15:07:14.81-35:04:59.6 -34.1±2.4-30.1±2.5 10.59±0.06 l.OOiO.ll 8.87±0.01 8.42±0.05 8.35±0.03 J150714.5-350500 3 7.39e-02 0.71 58
7833-1106-1 UCL 15:08:00.55-43:36:24.9 -21.8±1.4-16.5±1.4 9.93±0.03 0.87±0.06 8.20±0.01 7.86±0.04 7.71±0.02 J150759.9-433642 18 8.84e-02 0.88 ...
7833-2400-1 UCL 15:08:37.75-44:23:16.9 -17.9±2.1-17.6±2.0 10.91±0.07 0.80±0.11 9.33±0.03 8.95±0.02 8.81rb0.03 J150836.0-442325 20 1.73e-01 0.45 59
7833-2559-1 UCL 15:08:38.50-44:00:52.1 -23.5±2.1-24.2±1.9 10.61±0.05 0.70±0.09 8.95±0.02 8.54±0.03 8.54±0.03 J150838.5-440048 4 1.21e-01 0.19 60
8293-92-1 UCL 15:09:27.93-46:50:57.2 -19.9±2.1-16.1±2.0 10.60±0.05 1.09i:0.09 8.39±0.04 7.83±0.04 7.72±0.02 J150928.2-465109 12 7.77e-02 1.00 ...
8294-2230-1 UCL 15:12:50.18-45:08:04.5 -19.8±2.1-22.4±2.0 10.79±0.07 0.85it0.11 9.13±0.01 8.74±0.03 8.67±0.02 J151250.0-450822 17 1.45e-01 0.19 61
8694-1685-1 UCL 15:18:01.74-53:17:28.8 -28.7±2.5-28.3±2.3 10.21±0.04 0.86±0.07 8.51±0.02 8.12±0.04 8.01±0.02 J151802.0-531719 10 2.23e-01 -0.04 62
7822-158-1 UCL 15:18:26.91-37:38:02.1 -22.5±2.6-28.3±2.6 11.11±0.09 0.87±0.15 9.07±0.03 8.63±0.04 8.53±0.03 J151827.3-373808 7 1.30e-01 0.37 63
8295-1530-1 UCL 15:25:59.65-45:01:15.8 -21.9±2.2-21.3±2.0 10.98±0.07 0.88±0.12 9.45±0.02 8.98±0.02 8,88±0.02 J152600.9-450113 13 6.10e-02 -0.15 64
7326-928-1 UCL 15:29:38.58-35:46:51.3 -21.7±2.0-26.1±2.1 10.54±0.06 1.07±0.11 8.77±0.01 8.27±0.03 8.19±0.03 J152937.7-354656 11 1.05e-01 0.21 65
7318-593-1 UCL 15:31:21.93-33:29:39.5 -25.8±2.5-31.5±2.5 10.99±0.08 0.72±0.12 9.39±0.03 8.95±0.06 8.81±0.03 J153121.8-333002 23 9.34e-02 0.06 ...
7327-1934-1 UCL 15:37:02.14-31:36:39.8 -17.8±1.6-29.0±1.6 10.06±0.04 0.70±0.06 8.28±0.03 7.76±0.04 7.72±0.02 J153701.9-313647 7 3.92e-01 0.32 66
7840-1280-1 UCL 15:37:11.30-40:15:56.8 -16.7±1.9-22.5±1.9 10.55±0.05 1.24±0.10 8.98±0.01 8.46±0.05 8.31±0.03 J153711.6-401608 11 2.27e-01 -0,14 67
7848-1659-1 UCL 15:38:43.06-44:11:47.4 -23.2±1.6-28.4±1.5 10.36±0.05 0.85±0.08 8.81±0.05 8.36±0.04 8.22±0.03 J153843.1-441149 1 1.41e-01 -0.01 68
6785-510-1 UCL 15:39:24.41-27:10:21.9 -20.2±1.7-29.8±1.6 9.65±0.03 0.81±0.05 8.04±0.03 7.64±0.03 7.53±0.01 J153924.0-271035 14 2.33e-01 0.59 69
7331-782-1 UCL 15:44:03.77-33:11:11.2 -22.2±2.6-29.7±2.6 10.97±0.07 0.95±0.12 9.05±0.02 8.55±0.01 8.41±0.03 J154404.1-331120 10 7.72e-02 0.42 70
7845-1174-1 UCL 15:45:52.25-42:22:16.5 -18.0±1.5-30.4±1.6 10.61±0.05 1.15±0.09 8.69±0.02 8.09±0.04 7.94±0.02 J154552.7-422227 11 4.13e-01 0.07 71
8317-551-1 UCL 15:46:51.79-49:19:04.7 -19,7±1.6-27.9±1.5 10.29±0.04 1.04±0.08 8.69±0.01 8.36±0.03 8.31±0.02 J154651.5-491922 17 1.78e-01 0.07 72
#
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
Table 2.2—Continued
(2) (3) (4) (5) (6) (7) (8) (9) (10) (11)
OB «, 5 (ICRS) f ig VT (BT - Vt) J H Ks IRXS sep.
Grp. h m s ° " (masyr"^) (mag) (mag) (mag) (mag) (mag) Name (")
UCL 15:56:59.05-39:33:43.1 -15.1±2.0-21.6di2.0 10.90±0.07 0.94±0.12 9.20±0.02 8.83±0.03 8.70±0.03 J155659.0-393400 16
UCL 16:01:07.93-32:54:52.5 -18.3±1.4-29.6±1.5 9.58±0.02 0.73±0.04 8.46±0.01 8.16dz0.04 8.08±0.02 J160108.0-325455 3
UCL 16:01:08.97-33:20:14.2 -12.3±2.1-23.0±2.2 10.99±0.07 1.15±0.16 9.01±0.02 8.55±0.03 8.46±0.03 J160108.9-332021 7
UCL 16:03:45.37-43:55:49.2 -12.5±1.1-22.6±1.4 9.74±0.03 1.06±0.05 7.93±0.02 7.42±0.03 7.30±0.01 J160345.8-435544 6
UCL 16:03:52.50-39:39:01.3 -16.4±2.4-28.4ib2.3 11.12±0.08 1.15±0.18 8.98±0.03 8.35±0.03 8.21±0.03 J160352.0-393901 5
UCL 16:05:45.00-39:06:06.5 -16.8±2.2-31.1±2.2 10.63±0.06 l.OOiO.lO 8.93±0.02 8.52±0.04 8.36±0.03 J160545.8-390559 11
UCL 16:13:58.02-36:18:13.4 -18.4±2.0-32.3±2.1 11.26±0.09 1.17±0.21 9.41±0.02 8.98±0.03 8.84±0.03 J161357.9-361813 1
UCL 16:14:52.01-50:26:18.5 -19.9±2.1-29.2±2.0 10.50±0.05 0.96±0.09 8.38±0.01 7.91±0.01 7.76±0.01 J161451.3-502621 7
UCL 16:18:38.56-38:39:11.8 -25.9±1.3-34.0±1.5 9.09±0.02 0.69±0.03 8.02±0.01 7.79±0.05 7.69±0.02 J161839.0-383927 16
UCL 16:21:12.19-40:30:20.6 -9.1±1.4-28.1±1.4 10.67±0.05 1.01±0.09 8.97±0.02 8.55±0.02 8.43±0.02 J162112.0-403032 11
UCL 16:23:29.55-39:58:00.8 -12.5±2.1-24.1±2.1 10.73±0.06 0.88±0.09 9.47±0.02 9.10±0.03 8.99±0.03 J162330.1-395806 8
UCL 16:27:30.55-37:49:21.6 -10.0±2.5-23.4±2.6 11.05±0.07 0.84±0.12 9.29±0.02 8.81±0.04 8.66±0.03 J162730.0-374929 9
UCL 16:31:42,03-35:05:17.2 -16.2±1.9-27.6±2.0 10.72±0.05 0.81±0.08 9.19±0.02 8.75±0.02 8.63±0.02 J163143.7-350521 20
UCL 16:35:35.99-33:26:34.7 -6.0±1.8-23.0±2.1 11.09±0.07 0.93±0.13 8.99±0.02 8.44±0.03 8.28±0.02 J163533.9-332631 25
UCL 16:38:38.47-39:33:03.5 -10.4±1.3-19.2±1.3 9.05±0.02 0.73±0.04 7.52±0.01 7.25±0.02 7.15±0.01 J163839.2-393307 9
UCL 16:39:59.30-39:24:59.2 -9.7±2.2-18.9±2.2 10.68±0.08 0.84±0.12 8.67±0.02 8.26±0.04 8.12±0.02 J163958.7-392457 7
UCL 16:42:24.00-40:03:29.7 -11.1±1.5-20.2±1.3 9.70±0.03 0.88±0.06 8.03±0.03 7.63±0.05 7.46±0.02 J164224.5-400329 5
40
2.3 Observations
Blue and red optical spectra of the pre-MS candidates were taken simultaneously
with the Dual-Beam Spectrograph (DBS) on the Siding Springs 2.3-m telescope
on the nights of 20-24 April 2000. The DBS instrument is detailed in Rodgers,
Conroy, & Bloxham (1988). Using a 2" wide slit, we used the B6001/mm grating
in first order on the blue channel, yielding 2.8A FWHM resolution from 3838-
5423A. The red channel observations were done with the R12001/mm grating in
first order, yielding 1.3A FWHM resolution over 6205-7157A. The five nights of
bright time were predominantly clear to partly cloudy. Signal-to-noise ratios of
~50-200 per resolution element were typically reached with integration times of
120-720 s. Flat-fields and bias-frames were observed at the beginning and end
of each night. NeAr A-calibration arcs and spectrophotometric standards were
observed every few hours. The spectra were reduced using standard IRAF rou
tines. In order to remove low-order chromatic effects from the band-ratio mea
surements, we spectrophotometrically calibrated all of the target spectra using
2 standard stars from Hamuy et al. (1994). A total of 118 program stars (§2.2.1
and §2.2.2) and 20 MK spectral standards (§2.4.1.1) were observed. The major
stellar absorption features of one of the single standard G stars were shifted to
a zero-velocity wavelength scale. The spectra of all of the stars were then cross-
correlated against this standard star using the IRAF task/xcor, and then shifted to
the common, rest-frame wavelength scale. This was done to ensure proper iden
tification of weak lines, as well as to make sure that the band-ratio measurements
were sampling the same spectral range in each stellar spectrum.
Table 2.2—Continued
(1) (2) (3) (4) (5) (6)
Name OB a, S (ICRS) /Xq* t-^S VT {BT - Vr)
TYC Grp. h m s ° " (masyr ' ) (mag) (mag)
(7) (8) (9) (10) (11) (12) (13) (14)
J H Ks IRXS Sep. X-ray Hrdns. Table 5
(mag) (mag) (mag) Name (") (cts~^) HRl #
Note. — Columns; (1) Tycho-2 name, (2) OB subgroup region, (3) Tycho-2 position (ICRS, epoch 2000.0) (4) proper motion (a,
D components) in masyr~^(where Ha* = fiaCosS), (5) Vx magnitude, (6) {BT — VT) color, (7-9) 2MASS JHKs magnitudes, (10)
X-ray counterpart name in RASS-BSC, (11) optical-X-ray separation, (12) X-ray count rate (cts~^), (13) Hardness ratio HRl, (14)
Number in Table 5 if star is found to be "pre-MS" or "pre-MS?" in nature. X-ray data is from ROSAT AU-Sky Survey BSC (Voges et
al., 1999), astrometry and optical photometry are from Tycho-2 catalog (H0g et al., 2000a), and near-IR photometry is from 2MASS
(Cutrietal.,2003).
42
2.4 Analysis of Spectra
2.4.1 Spectral Types and Luminosity Classification
2.4.1.1 Standard Stars
We observed 20 spectral standards including dwarfs and subgiants (luminosity
classes IV and V) and a few giants (III). A summary of their properties is listed in
Table 2.3. To permit quantitative examination of trends in the strengths of spectral
features, as well as interpolation between spectral types, we adopt the numerical
subtype scaling of Keenan (1984; i.e. here listed as "SpT", where GO = 30, G2 =
31, KO = 34, etc.). All of the standard stars are classified on the MK system by
Keenan & McNeil (1989), except for HR 7061 (Garcia, 1989). Table 2.3 also lists
their spectral types as given in the Michigan Spectral Survey atlases of N. Houk
(e.g. 1978). The sample standard deviation of a linear fit between the Keenan
and Houk spectral types for dwarfs and subgiants, on Keenan's subtype scale,
is (T(SpT) = 0.6 subtypes. The ~0.6 subtype uncertainty probably represents the
best that can be done using visual spectral types determined by different authors.
The adopted spectral types are those of Keenan & McNeil's, however the lu
minosity classification was verified (and some times changed) by virtue of (1)
position of the stars on a color-magnitude diagram based on Hipparcos data, (2)
position in a temperature vs. Sr II A4077/Fe I A4071 ratio diagram (see §2.4.1.3,
Fig. 2.1), and (3) published log g estimates. Although changing the classifica
tion of some standards may appear imprudent, the H-R diagram positions, Sr/Fe
line ratio, and derived log y values (Cayrel de Strobel et ah, 2001) are all consistent with
our new adopted luminosity classes^. In every case the difference was only half of a
^Hipparcos data also led Keenan & Barnbaum (1999) to revise the luminosity classes of a few giant star standards at the half luminosity class level.
43
luminosity class, and only 5/20 of the stars were changed. Notes on the revised
luminosity classifications are given in Appendix B.
0.2
0.1
"HR 5553 (K0.5V)
0
6756 (KlIV)
• V dwarfs • IV subgiants A III giants X Hipparcos o RASS-ACT/TRC
HR 4287 (KO+III)
K4 K5 GO G5 KO K2 K3 K1 -0.3
0.2 0 0.1
MI6 [mag]
Figure 2.1: The MI6 band-ratio (T^jj indicator) vs. the band-ratio of Fe I A4071/
Sr II A4077 (surface gravity indicator). The solid line is a polynomial fit to only
the dwarf standards. The dashed lines separate dwarfs, subgiants, and giants.
The dwarf-subgiant dashed line is -2 x the a-residual below the dwarf regression,
whereas the subgiant-giant boundary is placed somewhat arbitrarily to resolve
the observed subgiant and giant loci. Empirically, this diagram suggests that
most of the target stars are consistent with being G and K-type subgiants, with
few giant and dwarf interlopers. A few early-K standards are noted for reference.
44
Table 2.3. Spectral Standard Stars
(1) (2) (3) (4) (5) (6) (7) (8) (9) (10)
Name Name MK V TT ± ( B - V ) [Fe/H] MSS Adopted notes
HD HR Sp.Type (mag) (mas) (mag) adopted Sp.Type Sp.Type
182640 7377 FOIV 3.36 65,1 ±0.8 0.32 F2V FOIV a.,6
173677 7061 F6V 4.19 52.4 ±0.7 0.48 -O.l F6IV-V c
84117 3862 F9V 4.93 67.2 ±0.7 0.53 GOV F9V
121370 5235 GOIV 2.68 88.2 ±0.8 0.58 0.2 GOIV b
89010 4030 G1.5IV-V 5,95 32.9 ±0.9 0.66 0.0 G1.5IV-V
126868 5409 G2IV 4.84 24.2 ±1,0 0.69 0.0 G3V G2in-IV &,d
161239 6608 G2inb 5.73 26.1 ±0.6 0.68 G2IV
146233 6060 G2Va 5.49 71.3 ±0.9 0.65 0.0 G5V G2V a
94481 4255 G4in 5.65 8.0 ±0.8 0.83 KOni+(G) G4ni
117176 5072 G4V 4.97 55.2 ±0.7 0.71 -0.1 G4IV-V e
188376 7597 G5IV 4.70 42.0 ±0,9 0.75 -0,1 G3/5in G5rv
115617 5019 G6.5V 4.74 117.3 ±0.7 0.71 0.0 G5V G6.5V
114946 4995 G7IV-V 5.31 25.9 ±0.7 0.86 -0.1: G8in/IV G7IV
188512 7602 G8IV 3.71 73.0 ±0.8 0.86 0.0: Gsrv a
165760 6770 G8in 4.64 13.7 ±0.8 0.95 -0.1 GSin
95272 4287 KO+ni 4.08 18.7 ±1.0 1.08 -0.1: Klffl Ko+m
131511 5553 K0.5V 6.00 86.7 ±0.8 0.84 K0.5V a
165438 6756 KlIV 5.74 28.6 ±0.8 0.97 0.0 KOIV KlIV
131977 5568 K4V 5.72 169.3 ±1.7 1.02 0.0 K4V K4V
120467 K6Va 8.16 70.5 ± 1.0 1.26 K6V
Note. — SB1,2 = spectroscopic binary in either SIMBAD or Duquennoy & Mayor
(1991). MSS = Michigan Spectral Survey Vols. 1-5 = Hotik & Cowley (1975), Houk (1978),
Houk (1982), Houk & Smith-Moore (1988), Houk & Swift (1999). The [Fe/H] estimate is
adopted from the compilation of published values in Cayrel de Strobel et al. (2001). A
semi-colon after the [Fe/H] value indicates considerable scatter (>0.2 dex) in the pub
lished estimates.
''SIMBAD lists as variable or suspected var., however Hipparcos finds scatter in Hp <
0.015 mag. The typical scatter for the other standard stars was 0.005 mag in Hp (Hippar
cos magnitude), with none greater than 0.01 mag.
''Spectroscopic binary
®From standard list of Garcia (1989), and originally in Johnson & Morgan (1953), but
not listed in either Keenan & Yorka (1988) or Keeiwi & McNeil (1989).
''HR 5409 is a resolved binary (sep. 5") listed by SIMBAD as a variable star, however
Hipparcos found the scatter in the Hp band to be only 0.015 mag.
®Keenan & McNeil (1989) call it G4V, however it is G5V in virtually every other refer
ence Grav. Graham. & Hovt. 20011. We retain Keenan's classification.
45
2.4.1.2 Visual Classification
The blue spectrum of each star was assigned a spectral type visually by E.M.
through comparison with the standards in Table 2.3. In order to distinguish sub
types, we focused on several features such as the G band (A4310), Ca I A4227,
Cr I A4254 and nearby Fe lines, and the Mg b lines A5167, A5173, A5184. Balmer
lines were ignored due to possible chromospheric emission. After making an ini
tial guess through comparison with a wide range of spectral types, a final visual
spectral type was assigned through comparison to standards within ±2 subtypes
of the initial guess.
To test the accuracy of our visual classification, we compared our spectral
types to those of quality 1 or 2 in the Michigan Spectral Survey. The average dif
ference is not significant; -0.4 ± 0.6 (la sample standard deviation) subtypes (on
Keenan's scale). The 12 Hipparcos stars were later visually typed a second time.
Between the two estimates for each star, we estimate that the la uncertainty in
our visual spectral types is 0.6 subtypes. This is comparable to the dispersion
between the Keenan and Houk spectral types for the standards themselves.
2.4.1.3 Quantitative Spectral Type Estimation
A two-dimensional quantitative spectral type (subtype plus luminosity class) can
be estimated using integrated fluxes over narrow bands sensitive to temperature
and surface gravity. We tested various ratios defined by Malyuto & Schmidt-
Kaler (1997) and Rose (1984) for this purpose, as well as from Gray's spectral
atlas (2000). In testing band-ratios as temperature indicators for our standards,
we noticed that some had slight surface gravity dependencies. A surface grav
ity dependence in our temperature indicators could systematically affect our Teff
46
estimates. We first discuss our surface gravity indicator and then define our tem
perature estimators using only subgiant and dwarf standards, thus mitigating the
effects of surface gravity.
The most widely used surface gravity diagnostic for G and K stars is the ratio
between Sr IIA4077 and nearby Fe lines (e.g. Keenan & McNeil, 1976; Gray, 2000).
In thin and thick disk dwarfs, the abundance ratio [Sr/Fe] is within ~0.1 dex of
solar for most stars (Mashonkina & Gehren, 2001). A quantitative surface gravity
(luminosity class) indicator was established by Rose (1984) from low-resolution
spectra using the maximum absorption line depth for Sr IIA4077 and the average
for the atomic Fe A4045 and A4063 lines. We measure the fluxes in 3 A bands cen
tered on the Sr IIA4077 line and the Fe IA4071 line. Ratios between the A4077 line
and the other nearby Fe lines (A4045, A4063) did not distinguish subgiants and
giants.
For a temperature estimator, we adopted Index 6 of Malyuto & Schmidt-
Kaler (1997) (AA5125-5245/AA5245-5290) hereafter referred to as "MI6". The tem
perature sensitivity of this indicator largely reflects differing amounts of line-
blanketing in these two wavelength regimes - mainly by the Mg b lines (A5167,
A5173, and A5184), and many Fe lines (e.g. Fe I A5270). Although the Mg b lines
are somewhat surface gravity sensitive, within the log g and Tg/f regime of our
standards and program stars, the temperature sensitivity is dominant. The dif
ference in central wavelength between the two bandpasses is only 82A, and the
effects of reddening are negligible (Mathis, 1990). For a temperature indicator, we
fit a low-order polynomial to MI6 vs. SpT for the dwarf and subgiant standard
stars (see Appendix C) which has a la sample standard deviation of 0.6 subtypes.
47
Fig. 2.1 plots the temperature-sensitive MI6 index versus our surface gravity
discriminant (Sr II A4077/Fe I A4071). The dwarf standards form a very narrow
sequence in Fig. 2.1, confirming the lack of cosmic scatter in [Sr/Fe] values among field
stars and the insensitivity of log g to spectral type for G-K dwarfs. The polynomial fit
to the dwarf data is given in Appendix C. There is a gap between the dwarf and
subgiant loci between ~1-2.5(T (sample standard deviation) of the dwarf locus
polynomial, and we set the subgiant/dwarf separation at la. We classify stars
within 2(7 of the solid dwarf line in Fig. 2.1 as dwarfs (4 of 96 RASS-ACT/TRC
stars, 4 of 20 HIP stars), and three stars near the giant locus (TYC 8992-605-1, HIP
68726, and HIP 74501) as giants. We classify the rest as subgiants.
Gray (2000) suggests Y II A4376/Fe I A4383 as a surface gravity indicator for
late-G stars using low-resolution spectra. From the solar spectral atlas of Wal
lace, Hinkle, & Livingston (1998), it appears that Gray's low-resolution YIIA4376
feature is actually a blend of several lines of nearly equal strength. In order to
test the properties of this band-ratio, we measure the flux in 3A windows cen
tered on wavelengths 4383.6A and 4374.5A. Plotting this ratio against spectral
type for the standard stars showed a very tight locus for the dwarfs, however
luminosity classes IV-V, IV, and III were indistinguishable from the dwarfs and
each other. We found this ratio unsuitable for the purposes of luminosity classi
fication of our targets, but we find it to be an excellent temperature estimator for
FGK dwarfs, subgiants, and giants. Among the 20 standards, the measurement
of the A4374/A4383 band-ratio vs. spectral type gives a tight correlation (sample
standard deviation la = 0.6 subtypes). We adopt the A4374/A4383 band ratio as
our third, independent estimator of spectral type (polynomial fit is given in Ap
pendix C).
48
2.4.1.4 Final Spectral Types
The three temperature-type estimates agree well for the majority of the program
stars. The mean difference between the MI6 and visual spectral types is 0.7 sub
types. The mean difference between the A4374/A4383 band-ratio types and the
visual types is 0.6 subtypes. We calculate a mean spectral type and standard er
ror of the mean using the three classifications. The mean is unweighted since all
three relations appeared to have la sample standard deviations of wO.6 subtypes
in their accuracy. The average standard error of the mean is 0.5 subtypes. We
believe that using multiple techniques mitigates the effects that rapid rotation,
binarity, etc. can introduce into visual classification alone. The spectral types are
listed in Tables 2.6,2.5 and 2.7.
2.4.2 Additional Spectroscopic Diagnostics
2.4.2.1 Chromospheric Hct Emission
Medium-to-low resolution spectra of chromospherically active stars show the HO
line to be partially filled-in, or even fully in emission. We measure the EW of
the entire HQ feature; our resolution is insufficient to separate the "core" chro
mospheric emission from the photospheric Ha absorption line. A significant
number of our stars show HA emission (19% of the RASS-ACT/TRC G-K type
stars). We characterize our targets stars as chromospherically active or inactive
through comparing the Ha EW to that of standards of identical spectral type. Fig.
2.2 shows the EW(HQ) data for our targets and standard stars. Stars more than
2a above the dwarf/subgiant EW(HQ) relation (a quadratic regression; see Ap
pendix C) are considered to be active. The stars with Ha in emission (negative
49
EWs) have an "e" appended to their spectral types in Tables 2.6,2.5 and 2.7. The
Htt EWs for each star are also listed in these tables.
- 2
0
Xoo X
2
®V dwarfs • IV subgiants X Hipparcos o RASS-ACT/TRC
KO G5 GO K2 4
32
Spectral Type
34 36 28 30
Figure 2.2: Ho EWs for the pre-MS candidates compared to inactive field dwarfs
and subgiants. Symbols are the same as for Fig. 2.1. The solid line is the average
EW(Ha) for dwarf and subgiant standard stars. The dashed line represents the
±2cr residual scatter in the relation (encompassing all of the standards). Stars
above this line are clearly chromospherically active, however those within the 2a
scatter have Ha emission similar to older field stars.
2.4.2.2 Li I A6707 Equivalent Width
The presense of strong Li absorption in the spectra of late-type stars is a well-
known diagnostic of stellar youth. Because of the extended timescale for signifi
cant Li depletion in stars of ~1 M©, strong Li absorption is necessary, but not sufficient
50
indicator of pre-MS nature for G stars. However it is a powerful age descriminent
when combined with our surface gravity indicator.
Many studies have shown that the equivalent width (EW) of Li I A6707 can
be overestimated at low spectral resolution (e.g. Covino et al., 1997), especially
for G-K stars. With a resolution of 1.3A, we consider our EWs to be approximate.
The EWs were measured with Voigt profiles in the IRAF routine splot. The contin
uum level was estimated from nearby pseudo-continuum peaks. We subtract the
contribution from the neighboring Fe I A6707.4A feature using the prescription
of Soderblom et al. (1993). In order to test the validity of our Li EWs, we divided
several of our Li-rich targets by standard stars of the same spectral type. The ra-
tioed spectra exhibit only a major absorption feature at A6707. The division also
removes the effects of blending by Fe lines (assuming the same stars have similar
EWs). The EWs of this feature in the divided spectra corresponded well with our
previous measurements, however the uncertainties in the EW appear to be ~20-
50 mA (with the maximum value being for spectroscopic binaries).
Fig. 2.3 shows the Li I A6707 EWs for our RASS-ACT/TRC and Hipparcos
targets, separated according to their luminosity class (§2.4.1.3). Effective temper
atures (Teff) come from the final spectral type (see §2.6.2). Most points lie above
the Li I A6707 EWs that characterize young open clusters, plotted as low-order
polynomial fits for the IC 2602 (30 Myr; Randich et al., 1997), Pleiades (70-125
Myr; Soderblom et al., 1993; Basri, Marcy, & Graham, 1996), and M34 clusters (250
Myr; Jones et al., 1997). The comparison is not completely fair, however, since the
cluster ZAMS stars will be roughly 10% less massive than the corresponding pre-
MS stars. Even if most of our program stars were older ZAMS stars, they still
51
would be Li-rich compared to stars in the well-studied open clusters. We select
as "Li-rich" those stars above the solid line in Fig. 2.3. Considering the uncer
tainties in our EW(Li) measurements, and the lack of any other ~ 10-20 Myr-old
pre-MS G-K-type stellar samples with which to compare, we are not compelled
to subdivide our sample further. For the present, we are content to have demon
strated that we have identified a population which appears to be more Li-rich
than ZAMS stars.
G5 KO K2 GO 400
• •
• « • IC 2602 (~30 Myr)
%• Pleiades 120 Myr)
M34 (~250 Myr)
100
O Dwarf • Subgiant
0 3.7 3.75
log(T,„) [K]
Figure 2.3: EWs for Li I A6707 for the program stars compared to regression fits
for stars in young open clusters (see §2.4.2.2). We discuss assignment of dwarf
and subgiant luminosity classes in §2.4.1.3. The pre-MS candidates form an obvi
ous locus, and we select all stars above the solid line as "Li-rich".
52
2.5 Defining the PTTS Sample
2.5.1 Membership Status
Our survey was designed to identify the pre-MS G and K-type stars in the Sco-
Cen OB association. We classified the late-type stars according to their positions
in Figs. 2.1/ 2.2 and 2.3 (Table 2.4). We consider the 110 stars (85/96 RASS-
ACT/TRC and 16/30 Hipparcos) classified as "Li-rich", "subgiant", and "active"
as bona fide post-T Tauri stars ("pre-MS"). Li-rich stars with subgiant surface
gravities and Ha EWs similar to the standard field stars (i.e. "inactive") are
called "pre-MS?". Only 3 of the RASS-ACT/TRC stars, and 6 of the HIP stars
are classified as "pre-MS?". The lone object with giant-like surface gravity in
the RASS-ACT /TRC X-ray-selected sample (TYC 8992-605-1) is Li-rich, and we
also classify it as a pre-MS PTTS. The 9 "pre-MS?" stars were included in our
statistics concerning the star-formation history and disk-frequency of the sample
(§2.7) for a total of 110 candidate lower-mass members of the LCC and UCL sub
groups. All 13 stars selected in the RASS-ACT /TRC sample which overlaped
with Z99's membership lists were found to be pre-MS candidates. Our RASS-
ACT /TRC sample (including the 13 Z99 stars also selected) yielded a pre-MS
hit-rate of (88+13)/(96+13) = 93%. Of the 30 Z99 candidates we observed, 22/30
(73%) are classified as pre-MS or "pre-MS?". The numbers of stars by member
ship class are listed in Table 2.4. Pre-MS stars in the Hipparcos sample are listed
in Table 2.5, and those in the RASS-ACT/TRC sample are given in Table 2.6. Li-
rich stars with dwarf-like surface gravity (N = 5) were considered young main
sequence field stars ("ZAMS"), and are listed along with other interlopers in Ta
ble 2.7.
53
2.5.2 Sample Contamination
The primary contaminants one would expect from an X-ray- and proper motion-
selected sample are X-ray-luminous ZAMS stars (ages « 0.1-1 Gyr). Field ZAMS
stars could occupy the same region of UVW velocity space, and be selected in
our study by virtue of their proper motions and X-ray emission. However, our
selection of candidate Sco-Cen members utilizes a surface gravity criterion which
should minimize contamination. Even if our surface gravity indicator was in er
ror, we claim ZAMS stars do not dominate our sample. Field ZAMS stars exhibit
a large spread in Li EWs (especially for the late-G and early-K stars), however this
is not observed in Fig. 2.3. The star just below the "Li-rich" line in Fig. 2.3 (TYC
7318-593-1; G9, EW(Li) ~ 150 mA) happens to be the sole RASS-ACT/TRC star
with inferred log g higher than that of the dwarf standards in Fig. 2.1. We con
sider TYC 7318-593-1 to be a field ZAMS star candidate due to its intermediate Li
strength and high surface gravity.
54
Table 2.4. Stellar Classification Scheme
(1) (2) (3) (4) (5) (6)
Li- Ha Lum. N(R-T) N(HIP) Adopted
Rich Excess Class # # Class.
Yes Yes IV 94 7 Pre-main sequence (PMS)
Yes No IV 3 6 Probably pre-main sequence? (PMS?)
Yes Strong IV 1 0 Pre-MS Classical T Tauri star (CTTS)'^
Yes Yes/No V 4 1 Young Dwarf (ZAMS)
No Yes V 1 0 Active Dwarf
No Yes IV 4 0 Active Subgiant
No No IV 1 2 Subgiant
No No III 0 2 Giant
No No V 0 2 Dwarf (MS)
No No(Wide) IV 2 0 Chromosph.-Active Binary (CAB)*^
107 20 Total
Note. — N(R-T) is number in RASS-ACT/TRC sample. N(HIP) is number in
Hipparcos sample. "Li-rich" implies significant Li absorption, and above the line
in Fig.2.3. "Active" means that the Ho equivalent widths are >2 x the a-residual
above the regression of values for field, standard stars (Fig. 2.2; Appendix C),
implying that chromospheric emission is filling in the absorption line. Luminos-
ity classes are assigned according to a star's placement in Fig. 2.1.
'^included in Pre-MS sample count.
included in subgiant sample count.
Table 2.5. LCC & UCL Pre-Main Sequence Members from de Zeeuw et al.
(1999) List
(1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11) (12) (13) (14) (15) (16) (17) (18) (19)
Name OB a, S (ICRS) 'TiHlP '^sec Pi Spec. log(Te//) logL/Lg Ay EW(Ha) EW(Li) log(Z,x) log(Lx/ DM97 PSOl SDFOO Class Notes
HIP Grp h m s ° " (mas) (mas) % Type (K) (dex) (mag) (A) (A) (ergs 1) ^bol) T ,M T ,M
57524 LCC 11:47:24.55 -49:53:03.0 9.62(1.39) 8.83(0.60) 81 F9IV 3.783(4) 0.44(6) 0.30(10) 2.0 0.18 30.8 -3.3 161.3 161.2 201.3 PMS
58996 LCC 12:05:47.48 -51:00:12.1 9.78(1.16) 9.79(0.52) 91 Girv 3.772(5) 0.43(5) 0.14(4) 3.3 0.21 30.5 -3.6 141.3 13 1.3 18 1.3 PMS?
59854 LCC 12:16:27.84 -50:08:35.8 8.24(1.78) 7.67(0.61) 99 GlIV 3.772(6) 0.49(7) 0.37(11) 1.6 0.20 30.9 -3.2 13 1.4 121.3 171.3 PMS a
60885 LCC 12:28:40.05 -55:27:19.3 7.06(1.13) 9.49(0.52) 14 GOIV 3.778(5) 0.46(5) 0.16(5) 2.7 0.13 30.3 -3.8 14 1.3 13 1.3 18 1.3 PMS?
60913 LCC 12:29:02.25 -64:55:00.6 10.48(1.12) 9.82(0.58) 100 G4.5IV 3.758(8) 0.43(5) 0.21(4) 2.0 0.19 <29.7 <-4.3 11 1,4 121.3 161.3 PMS?
62445 LCC 12:47:51.87-51:26:38.1 6.61(1.54) 7.70(0.55) 13 G4.5IVe 3.758(13) 0.62(6) 0.66(6) -1.8 0.24 30.7 -3.4 71.6 51.6 111.5 PMS SB3?
63847 LCC 13:05:05.29 -64:13:55.3 9.56(1.35) 10.18(0.61) 48 G3IV 3.764(7) 0.36(6) 0.33(6) 1.6 0.22 <29.7 <-4.3 141.3 151.2 191.3 PMS
65517 LCC 13:25:47.83 -48:14:57.9 9.59(1.44) 10.56(0.57) 50 G1.5IV 3.770(6) 0.02(5) 0.14(8) 1.9 0.17 30.4 -3.3 .. 1.0 391.0 ...1.0 PMS b
66001 LCC 13:31:53.61 -51:13:33.1 5.99(1.71) 7.88(0.62) 82 G2.5IV 3.766(11) 0.33(7) 0.11(10) 1,2 0.28 30.6 -3.2 15 1.2 181.2 21 1.2 PMS
66941 LCC 13:43:08.69 -69:07:39.5 8.05(0.95) 9.37(0.52) 97 G0.5IV 3.775(8) 1.09(5) 0.38(4) 2.1 0.16 31.1 -3.5 3 2.4 4 2.0 5 2.1 PMS c
67522 UCL 13:50:06.28 -40:50:08.8 7.94(1.64) 7.91(0.60) 93 G0.5IV 3.775(6) 0.25(7) 0.02(10) 2,2 0.15 30.4 -3.5 201.2 251.1 291.2 PMS
71178 UCL 14:33:25.78 -34:32:37.6 9.76(1.90) 8.71(0.56) 99 G8IVe 3.741(12) 0.16(6) 0.35(8) -0.2 0.26 <29.8 <-3.9 141.2 181.1 231.1 PMS
75924 UCL 15:30:26.29 -32:18:11.6 10.91(3.21) 9.79(0.79) 16 G2.5IV 3.765(6) 0.69(7) 0.37(7) 1,3 0.24 30.9 -2.9 71.6 51.7 111.5 PMS d
76472 UCL 15:37:04.66 -40:09:22.1 5.70(1.63) 7.24(0.49) 95 GlIV 3.774(8) 0.60(6) 0.29(10) 0,4 0.24 30.8 -3.3 11 1.4 111.4 141.4 PMS
77081 UCL 15:44:21.05 -33:18:55.0 6.63(1.61) 7.66(0.59) 69 G7.5IV 3.742(8) 0.35(7) 0.15(6) 1.8 0.25 <29.9 <-4.1 91.4 111.3 151.3 PMS?
77135 UCL 15:44:57.69 -34:11:53.7 6.98(2.92) 7.20(0.96) 91 G4IV 3.760(6) 0.42(12) 0.08(5) 2.2 0.23 30.5 -3.4 121.4 121.3 171.3 PMS? e
77144 UCL 15:45:01.83 -40:50:31.0 6.45(1.42) 8.00(0.53) 43 GOIV 3.778(7) 0.36(6) 0.08(5) 1.8 0.17 30.6 -3.4 171.2 201.2 23 1.3 PMS
77524 UCL 15:49:44.98 -39:25:09.1 6.63(1.93) 6.63(0.76) 96 Kl-rVe 3.705(13) 0.23(10) -0.03(10) -0.2 0.32 30.5 -3.3 41.4 9 1.3 101.4 PMS f
Table 2.5—Continued
(1) (2) (3) (4) (5) (6) (7) (8) (9)
Name OB a, 5 (ICRS) tth/p T^sec Pi Spec. log(Te//) logL/LQ
HIP Grp h m s ° > " (mas) (mas) % Type (K) (dex)
(10) (11) (12) (13) (14) (15) (16) (17) (18) (19)
Av EW(Ha) EW(Li) log(Lx) log(Lx/ DM97 PSOl SDFOO Class Notes
(mag) (A) (A) (e rgs - i ) L I O I ) T , M T , M T , M
3.756(9) 0.41(6) 0.52(9) 1.7 0.30 30.3 -3.8 111.4 12 1.3 16 1.3 PMS SB2
3.774(7) 0.65(7) 0.39(8) 1.5 0.21 30.9 -3.3 10 1.5 10 1.4 13 1.5 PMS ...
3.777(9) 0.74(10) 0.33(9) 2.5 0.14 <30.3 <-4.0 91.6 51.7 111.5 PMS ...
3.775(5) 0.78(8) 0.03(7) 2.8 0.13 <30.2 <-4.3 81.6 51.7 101.6 PMS? ...
Note. — Columns: (1) Hipparcos catalog ID, (2) OB subgroup, (3) Hipparcos position (ICRS, epoch 2000.0), (4) Hipparcos astrometric parallax (mas), (5) Secular parallax
estimate (mas) (see §2.6.3), (6) Membership probability (with VcU^p = 1 kms~^), (7) Spectral type (see §2.4.1), (8) Effective temperature (see §2.6.2), (9) Luminosity (see
§2.6.5), (10) Extinction Ay (see §2.6.5), (11) Equivalent Width of Ho A6562.8 (A), (12) Corrected EW of Li I A6707.8 (A), (13) X-ray luminosity (ergs~i) (see §2.2.2;
approximate upper limits assvune RASS PSPC detection limit of 0.05 ct s~ and HRl = 0), (14) Logarithm of ratio between X-ray and bolometric liuninosities, (15) Age
T ( in Myr) and Mass A I ( in M/M©) us ing DM97 t racks , (16) Age T ( inMyr) and Mass A 4 ( in M / MQ ) us ing PSOl t racks , (17) Age T ( in Myr) and Mass A 4 ( in M / MQ )
using SDFOO tracks, (18) Class - PMS = pre-main sequence, PMS? = pre-main sequence? (see §2.5.1), (19) Notes - SB = spectroscopic binary.
''HIP 59854 = CCDM J12165-5009AB. Unresolved binary (p = 0.2").
•'HIP 65517. The Michigan Spectral Survey (Houk, 1978) classifies this star as K0/2(V)+(G) (quality = 3). We foimd it to be much earlier (G1.5), and note that the
published (6 — y) and (B — V')colors support an early G classification.
' HTP 66941 = HD 119022 = Star "E" of Soderblom et al.'s (1998) high resolution survey of active southern stars. It is the most Li-rich of the very active stars in the
Soderblom sample. We did not resolve this tight, equal-brightness binary (p = 0.2"; CCDM J13431-6908AB).
dHIP 75924 = CCDM J15304-3218AB. Both stars were on the sUt (p = 1.5").
"HIP 77135 = [KWS97] Lupus 1 26 = CCDM J15450-3412AB. Binary is resolved (p = 3.4") but both were on-sUt. CJ1
ffflP 77524 = [KWS97] TTS 79 = CCDM J15450-3412AB. Unresolved binary (p = 0.2").
Table 2.6. Pre-Main Sequence LCC & UCL RASS-ACT/TRC Members
(1) (2) (3) (4)
ID Name OB a, S (ICRS)
# TYC Grp h m s ° '
(5) (6) (7) (8) (9) (10) (11) (12) (13) (14) (15) (16) (17) (18) (19)
Wsea Pi Spec. Iog(Te//) logL/Lo Ay EW(Ha) EW(Li) Iog(ix) log(Lx / DM97 PSOl SDFOO Class Notes
(mas) % Type (K) (dex) (mag) (A) (A) (e rgs""^) L to j ) T , M T , M
1 9212-2011-1 LCC 10:57:49.38 -69:13:59.9 9.76(101) 79 Ki+rv 3.699(10) 0,02(9) 0.35(6) 0.7 0.34 30.3 -3.3 7 1.2 13 1.1 15 1.2 PMS ...
2 8625-388-1 LCC 11:32:08.34 -58:03:20.0 10.78(121) 59 G7IV 3.747(13) 0.19(10) 0.68(13) 0.4 0.45 29.9 -3.7 15 1.2 19 1.1 23 1.2 PMS ...
3 8982-3213-1 LCC 12:04:48.87 -64:09:55.4 8.36(127) 12 Girv 3.773(5) 0.44(13) 0.60(11) 1.5 0.21 30.8 -3.2 14 1.3 13 1.3 18 1.3 PMS ...
4 8640-2515-1 LCC 12:06:13.54 -57:02:16.8 6.48(106) 90 G4IV 3.760(7) 0.15(14) 1.45(6) 0.6 0.23 30.6 -3.1 21 1.1 25 1.1 28 1.2 PMS ...
5 8644-802-1 LCC 12:09:41.86 -58:54:45.0 9.00(114) 65 KOIVe 3.720(13) 0.20(11) 0.28(10) -0.1 0.33 30.5 -3.2 7 1.4 12 1.2 15 1.3 PMS ...
6 8644-340-1 LCC 12:11:31.43 -58:16:53.2 9.29(91) 95 G9IV 3.729(8) 0.05(9) 0.28(10) 0.3 0.31 30.5 -3.2 14 1.2 21 1.1 24 1.1 PMS ...
7 9231-1566-1 LCC 12:11:38.14 -71:10:36.0 10.15(68) 91 G3.5IV 3.762(5) 0.39(6) 0.48(8) 1.4 0.27 30.5 -3.5 13 1.3 13 1.3 18 1.3 PMS ...
8 8636-2515-1 LCC 12:12:35.75 -55:20:27.3 9.24(90) 50 KO+IV 3.717(7) 0.00(9) 0.48(12) 0.2 0.34 30.3 -3.2 11 1.2 19 1.1 23 1.1 PMS ...
9 8242-1324-1 LCC 12:14:34.09 -51:10:12.5 9.41(83) 82 G9rv 3.733(8) 0.01(8) -0.02(5) 0.2 0.33 30.4 -3.2 17 1.1 25 1.0 28 1.1 PMS ...
10 8637-2610-1 LCC 12:14:52.31 -55:47:03.6 9.72(131) 62 G6IV 3.750(7) 0.28(12) 0.87(12) 0.6 0.27 30.9 -2.9 13 1.3 15 1.2 20 1.2 PMS a
11 8645-1339-1 LCC 12:18:27.64 -59:43:12.9 10.42(132) 4 K1.51Ve 3.695(10) 0.03(11) 0.12(8) -0.3 0.30 30.3 -3.1 6 1.2 12 1.2 14 1.2 PMS ...
12 8641-2187-1 LCC 12:18:58.02 -57:37:19.2 9.46(74) 97 G9.5IV 3.724(13) 0.32(7) 0.49(14) 0.1 0.35 30.6 -3.2 6 1.5 10 1.3 12 1.4 PMS ...
13 8983-98-1 LCC 12:19:21.64 -64:54:10.4 9.74(113) 89 Kl-IV 3.707(10) 0.26(10) 0.25(5) 0.3 0.32 30.5 -3.2 4 1.5 8 1.3 10 1.4 PMS ...
14 8633-508-1 LCC 12:21:16.48 -53:17:44.9 9.31(78) 27 G1.5IV 3.770(7) 0.31(7) 0.44(7) 1.6 0.21 30.7 -3.3 17 1.2 20 1.2 23 1.2 PMS ...
15 8238-1462-1 LCC 12:21:55.65 -49:46:12.5 9.85(66) 86 G5.5IV 3.754(7) 0.08(6) 0.14(6) 1.5 0.28 30.2 -3.4 22 1.1 28 1.0 30 1.1 PMS ...
16 8234-2856-1 LCC 12:22:04.30 -48:41:24.9 7.96(81) 85 KOIVe 3.721(12) 0.10(9) 0.21(7) -0.1 0.34 30.4 -3.2 10 1.3 15 1.1 19 1.2 PMS ...
17 8633-28-1 LCC 12:22:33.23 -53:33:49.0 8.07(69) 53 GOIV 3.778(5) 0.41(8) 0.41(11) 2.2 0.20 30.3 -3.7 16 1.3 17 1.2 20 1.3 PMS ...
18 8641-1281-1 LCC 12:23:40.13 -56:16:32.5 8.92(119) 66 Ko+rv 3.715(17) -0.09(12) 0.44(9) 1.1 0.36 30.0 -3.4 14 1.1 24 1.0 28 1.0 PMS ...
19 8992-605-1 LCC 12:36:38.97 -63:44:43.5 9.73(82) 56 Ko+m 3.717(17) 0.25(8) 0.39(6) 0.6 0.37 30.4 -3.4 5 1.5 10 1.3 12 1.3 PMS?b
20 8646-166-1 LCC 12:36:58.97 -54:12:18.0 8.64(118) 92 KOIV 3.721(6) 0.04(12) 0.11(5) 0.7 0.32 30.3 -3.4 11 1.2 18 1.1 22 1.1 PMS ... Ul
Table 2.6—Continued
(1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11) (12) (13) (14) (15) (16) (17) (18) (19)
ID Name OB a, 5 (ICRS) ^AEC Pi Spec. log(^e//) logL/L0 Av EW(Ha) EW(Li) log(Lx) log(Lx/ DM97 PSOl SDFOO Class Notes
# TYC Grp h m s ° " (mas) % Type (K) (dex) (mag) (A) (A) (ergs 1) Lboi) T,A4 t,M. t,M . . .
21 8654-1115-1 LCC 12:39:37.96 -57:31:40.7 10.22(96) 3 G7.5IV 3.742(12) -0.06(8) 0.30(7) 0.1 0.26 30.4 -3.1 241.0 34 0.9 47 1.1 PMS
22 8659-2604-1 LCC 12:41:18.17 -58:25:56.0 10.26(95) 58 G1.5rVe 3.771(11) 0.14(8) 0.67(14) -0.6 0.29 30.2 -3.4 24 1.1 30 1.1 300 1.2 PMS
23 8992-420-1 LCC 12:44:34.82 -63:31:46.2 7.97(119) 59 KirVe 3.704(12) 0.25(13) 0.27(8) -0.7 0.38 30.6 -3.0 41.4 81.3 9 1.4 PMS c
24 8249-52-1 LCC 12:45:06.76 -47:42:58.2 8.34(62) 94 G8.5IV 3.738(11) 0.13(7) 0.38(14) 0.0 0.30 30.6 -3.1 14 1.2 19 1.1 23 1.1 PMS
25 7783-1908-1 LCC 12:48:07.79 -44:39:16.8 10.97(59) 21 G6IVe 3.751(13) 0.14(5) 0.18(8) -0.1 0.28 30.6 -3.1 18 1.1 23 1.1 26 1.1 PMS
26 8655-149-1 LCC 12:48:48.18 -56:35:37.8 7.56(73) 56 G5IV 3.755(8) 0.14(9) 0.53(13) 1.2 0.22 30.4 -3.4 20 1.1 24 1.1 271.1 PMS
27 9245-617-1 LCC 12:58:25.58 -70:28:49.2 11.73(69) 80 Ko+rv 3.714(10) -0.01(6) 0.46(12) 0.0 0.37 30.3 -3.3 111.2 181.1 22 1.1 PMS d
28 8648-446-1 LCC 13:01:50.70-53:04:58.3 9.22(151) 75 K2-IV 3.690(8) -0.37(14) 0.55(10) 0.2 0.33 30.0 -3.3 19 0.9 39 0.8 41 0.9 PMS
29 8652-1791-1 LCC 13:02:37.54 -54:59:36.8 6.40(81) 16 GlIV 3.772(6) 0.35(11) 0.38(5) 2.1 0.17 30.8 -3.1 16 1.2 18 1.2 211.3 PMS
30 8258-1878-1 LCC 13:06:40.12 -51:59:38.6 8.91(90) 94 KOrVe 3.722(19) -0.05(9) 0.17(9) -0.3 0.34 30.2 -3.4 15 1.1 24 1.0 28 1.0 PMS
31 8259-689-1 LCC 13:14:23.84 -50:54:01.9 7.70(81) 87 G5.5IV 3.754(10) 0.25(10) 0.27(9) 1.1 0.32 30.8 -2.9 15 1.2 18 1.1 22 1.2 PMS
32 8649-251-1 LCC 13:17:56.94 -53:17:56.2 5.99(121) 56 GlIV 3.774(8) 0.43(18) 0.22(11) 1.0 0.30 30.7 -3.2 14 1.3 141.3 19 1.3 PMS
33 8248-539-1 LCC 13:22:04.46 -45:03:23.1 7.16(67) 71 Gorv 3.779(6) 0.18(8) 0.22(8) 2.1 0.15 30.3 -3.6 26 1.2 30 1.1 ... 1.2 PMS
34 9246-971-1 LCC 13:22:07.54 -69:38:12.3 11.57(95) 82 KllVe 3.702(17) 0.00(8) 0.17(7) -39.9 0.37 30.2 -3.2 8 1.2 14 1.1 171.2 PMS e
35 8663-1375-1 LCC 13:34:20.26 -52:40:36.1 9.50(66) 100 GUV 3.774(7) 0.32(6) 0.45(10) 1.7 0.22 30.5 -3.4 18 1.2 211.2 24 1.3 PMS
36 7796-1788-1 UCL 13:37:57.29 ^1:34:41.9 10.23(52) 49 KOIV 3.723(7) 0.00(5) -0.07(5) 0.7 0.30 30,4 -3.2 14 1.2 211.0 25 1.1 PMS
37 8667-283-1 LCC 13:43:28.53 -54:36:43.5 11.70(75) 2 G2IV 3.769(7) 0.13(6) 0.28(12) 1.7 0.20 30.6 -3.2 24 1.1 301.1 200 1.2 PMS
38 8270-2015-1 UCL 13:47:50.55 -49:02:05.5 6.75(79) 73 G8IVe 3.740(11) 0.05(10) 0.45(13) -0.5 0.29 30.5 -3.2 18 1.1 25 1.0 28 1.1 PMS
39 8263-2453-1 UCL 13:52:47.80 -46:44:09.2 6.92(57) 74 GOIV 3.778(8) 0.44(7) 0.34(8) 2.4 0.18 30.8 -3.3 15 1.3 15 1.3 19 1.3 PMS
40 7811-2909-1 UCL 14:02:20.73 -41:44:50.8 7.69(71) 58 G9IV 3.733(13) 0.07(8) 0.90(12) 0.2 0.23 30.3 -3.3 14 1.2 211.1 25 1.1 PMS
00
Table 2.6—Continued
(1) (2) (3) (4)
ID Name OB A, S (ICRS)
# TYC Grp h m s ° '
(5) (6) (7) (8) (9) (10) (11) (12) (13) (14) (15) (16) (17) (18) (19)
TTscc Pi Spec. log(Te//) logL/Lo Ay EW(Ha) EW(Li) log(Lx) log(Lx/DM97 PSOl SDFOO Class Notes
(mas) % Type (K) (dex) (mag) (A) (A) (ergs""^) Lf,oi) T,M T,M T,M
41 7815-2029-1 UCL 14:09:03.58 -44:38:44.4 7.05(49) 9 F9IV 3.783(5) 0.50(6) 0.38(6) 3.0 0.11 30.7 -3.5 141.3 13 1.3 18 1.3 PMS?
42 9244-814-1 LCC 14:16:05.67 -69:17:36.0 8.15(92) 11 Girv 3.773(7) 0.19(10) 0.62(8) 2.2 0.13 30.1 -3.7 231.1 27 1.1 58 1.2 PMS
43 8282-516-1 UCL 14:27:05.56 -47:14:21.8 7.57(63) 89 G7IV 3.745(10) 0.06(8) 0.17(4) 1.1 0.26 30.4 -3.3 191.1 26 1.0 29 1.1 PMS
44 7817-622-1 UCL 14:28:09.30 -44:14:17.4 6.23(71) 52 G5.5IV 3.754(8) 0.54(10) 0.33(7) 1.4 0.22 30.6 -3.5 81.5 9 1.4 12 1.4 PMS
45 7813-224-1 UCL 14:28:19.38 -42:19:34.1 6.27(61) 87 G3.5IV 3.762(8) 0.33(9) 0.37(7) 1.0 0.21 30.8 -3.0 141.3 16 1.2 20 1.2 PMS
46 7814-1450-1 UCL 14:37:04.22 -41:45:03.0 6.41(70) 88 G0.5IV 3.775(8) 0.59(10) 0.18(6) 1.5 0.17 30.7 -3.4 111.4 11 1.4 15 1.4 PMS f
47 8683-242-1 UCL 14:37:50.23 -54:57:41.1 7.59(94) 20 KO+rv 3.715(15) 0.08(11) 0.30(6) 0.1 0.32 30.6 -3.0 81.3 14 1.1 18 1.2 PMS
48 8283-264-1 UCL 14:41:35.00 -47:00:28.8 9.11(52) 94 G4IV 3.759(13) 0.20(6) 0.02(7) 1.2 0.23 30.9 -2.8 181.2 22 1.1 25 1.2 PMS
49 8283-2795-1 UCL 14:47:31.77 -48:00:05.7 7.68(81) 1 G2.5rV 3.766(8) -0.05(9) 0.43(7) 1.8 0.17 30.2 -3.4 44 1.0 1.0 PMS
50 7305-380-1 UCL 14:50:25.81 -35:06:48.6 5.27(87) 73 KOIV 3.723(15) 0.52(15) 0.21(8) 0.0 0.28 30.6 -3.3 31.8 5 1.6 7 1.7 PMS g
51 7828-2913-1 UCL 14:52:41.98 ^1:41:55.2 6.89(75) 84 KllVe 3.701(10) 0.10(10) 0.61(10) -0.6 0.35 30.5 -3.2 51.3 11 1.2 13 1.3 PMS
52 7310-2431-1 UCL 14:57:19.63 -36:12:27.4 7.59(61) 13 G6IV 3.750(11) 0.18(7) 0.15(6) 1.4 0.25 30.3 -3.4 161.2 20 1.1 24 1.2 PMS g
53 7310-503-1 UCL 14:58:37.70 -35:40:30.4 7.36(77) 82 K2-rVe 3.689(8) 0.21(9) 0.23(11) -0.3 0.37 30.3 -3.4 31.3 1.4 7 1.5 PMS g
54 7824-1291-1 UCL 14:59:22.76 -40:13:12.1 8.17(57) 85 G3IV 3.764(8) 0.32(6) 0.40(6) 1.5 0.23 30.5 -3.4 15 1.2 17 1.2 21 1.2 PMS g
55 7833-2037-1 UCL 15:00:51.88 -43:31:21.0 6.14(72) 78 G9IV 3.731(12) 0.16(11) 0.30(9) 0.1 0.33 30.4 -3.2 111.3 15 1.1 20 1.2 PMS g
56 7829-504-1 UCL 15:01:11.56 -41:20:40.6 4.67(62) 94 G0.5IV 3.774(8) 0.64(12) 0.48(8) 1.5 0.24 31.0 -3.3 101.5 10 1.4 13 1.4 PMS g
57 8297-1613-1 UCL 15:01:58.82 -47:55:46.4 6.43(62) 25 G1.5IV 3.770(6) 0.29(9) 0.22(10) 2.0 0.20 30.2 -3.8 18 1.2 21 1.2 24 1.2 PMS
58 7319-749-1 UCL 15:07:14.81 -35:04:59.6 9.92(82) 12 G9.5IV 3.727(11) -0.12(8) 0.44(10) 1,2 0.25 30.0 -3.4 211.0 33 0.9 38 1.0 PMS h
59 7833-2400-1 UCL 15:08:37.75 -44:23:16.9 5.55(72) 81 Gl.SIVe 3.770(11) 0.29(12) 0.13(11) -0.8 0.27 30.9 -2.9 181.2 21 1.2 24 1.2 PMS h
60 7833-2559-1 UCL 15:08:38.50 -44:00:52.1 7.46(71) 85 G1.5IV 3.771(9) 0.20(9) 0.11(7) 1.2 0.24 30.4 -3.3 211.1 26 1.1 30 1.2 PMS h Ol VD
Table 2.6—Continued
(1) (2) (3) (4) (5) (6) (7) (8) (9)
ID Name OB a, 5 (ICRS) iteec Pi Spec. log(Te/^) logL/L0
# TYC Grp h m s " ' " (mas) % Type (K) (dex)
(10) (11) (12) (13) (14) (15) (16) (17) (18) (19)
Av EW(Ha) EW(Li) log(Lx) log(Ljc/ DM97 PSOl SDFOO Class Notes
(mag) (A) (A) (ergs~i) Lj^;) T,M T,M T,M
61 8294-2230-1 UCL 15:12:50.18 -45:08:04.5 6.61(72) 100 G5IV 3.756(8) 0.19(10) 0.27(7) 1.7 0.24 30.6 -3.1 18 1.2 22 1.1 25 1.2 PMS h
62 8694-1685-1 UCL 15:18:01.74 -53:17:28.8 9.74(93) 14 G7IV 3.746(9) 0.08(9) 0.77(7) 1.2 0.27 30.4 -3.3 19 1.1 25 1.0 28 1.1 PMS
63 7822-158-1 UCL 15:18:26.91 -37:38:02.1 7.89(86) 97 G9IV 3.729(16) 0.05(10) 0.72(8) 0.4 0.26 30.4 -3.0 14 1.2 21 1.1 24 1.1 PMS h
64 8295-1530-1 UCL 15:25:59.65 -45:01:15.8 6.73(73) 40 G8IV 3.740(12) -0.02(10) 0.19(5) 1.1 0.26 30.1 -3.5 21 1.0 30 1.0 35 1.1 PMS h
65 7326-928-1 UCL 15:29:38.58 -35:46:51.3 7.39(71) 93 Ko+rv 3.718(10) 0.17(9) 0.16(5) 0.1 0.25 30.3 -3.4 7 1.4 12 1.2 15 1.2 PMS h
66 7327-1934-1 UCL 15:37:02.14 -31:36:39.8 7.41(59) 32 G6IV 3.750(9) 0.42(7) 0.43(5) 0.4 0.23 30.9 -3.0 9 1.4 11 1.3 15 1.3 PMS i
67 7840-1280-1 UCL 15:37:11.30 -40:15:56.8 6.11(67) 99 G8.5IVe 3.736(14) 0.25(10) 0.40(10) -0.9 0.30 30.7 -3.2 10 1.3 13 1.2 17 1.2 PMS h
68 7848-1659-1 UCL 15:38:43.06 -44:11:47.4 8.04(58) 82 G8.5IV 3.736(8) 0.08(7) 0.62(5) 0.5 0.27 30.3 -3.4 15 1.2 21 1.1 25 1.1 PMS h
69 6785-510-1 UCL 15:39:24.41 -27:10:21.9 7.89(61) 72 G5IV 3.755(7) 0.46(7) 0.28(6) 1.7 0.21 30.7 -3.3 9 1.4 11 1.3 15 1.4 PMS j
70 7331-782-1 UCL 15:44:03.77 -33:11:11.2 8.08(86) 99 KOIVe 3.723(13) 0.01(10) 0.47(6) -1.1 0.30 30.2 -3.3 13 1.2 21 1.1 25 1.1 PMS
71 7845-1174-1 UCL 15:45:52.25 -42:22:16.5 7.71(58) 32 K2-IVe 3.690(7) 0.12(7) 0.80(8) -0.8 0.39 30.9 -2.9 4 1.3 8 1.3 10 1.3 PMS h
72 8317-551-1 UCL 15:46:51.79 -49:19:04.7 7.55(59) 96 G7.5IV 3.744(7) 0.21(7) 0.29(8) 1.5 0.26 30.5 -3.3 13 1.2 17 1.1 21 1.2 PMS
73 7842-250-1 UCL 15:56:59.05 -39:33:43.1 5.74(70) 99 G9.5IV 3.726(7) 0.21(11) 0.35(6) 1.2 0.21 30.5 -3.4 8 1.4 12 1.2 16 1.2 PMS
74 7333-1260-1 UCL 16:01:07.93 -32:54:52.5 7.60(55) 87 GOIV 3.779(6) 0.33(6) 0.15(7) 2.3 0.11 30.6 -3.4 18 1.2 22 1.2 25 1.3 PMS
75 7333-719-1 UCL 16:01:08.97 -33:20:14.2 5.69(74) 64 G5IV 3.756(12) 0.41(12) 0.16(6) 0.7 0.26 30.6 -3.1 11 1.4 12 1.3 16 1.3 PMS h
76 7863-1629-1 UCL 16:03:45.37 -43:55:49.2 5.63(51) 62 G9.5IV 3.725(14) 0.75(8) 0.51(6) 0.7 0.35 31.2 -3.1 2 2.1 4 1.8 4 2.0 PMS h,k
77 7855-1106-1 UCL 16:03:52.50 -39:39:01.3 7.14(79) 83 K2-IVe 3.687(10) 0.09(10) 0.18(6) -0.4 0.38 30.4 -3.2 4 1.2 8 1.3 10 1.3 PMS h
78 7851-1-1 UCL 16:05:45.00 -39:06:06.5 7.69(75) 61 G6.5IV 3.749(10) 0.13(9) 0.19(5) 0.3 0.32 30.5 -3.1 18 1.1 23 1.1 26 LI PMS h
79 7355-317-1 UCL 16:13:58.02 -36:18:13.4 8.10(71) 90 G9IVe 3.732(14) -0.13(8) 0.31(9) -0.5 0.27 30.4 -2.9 24 1.0 37 0.9 45 1.0 PMS 1
80 8319-1687-1 UCL 16:14:52.01 -50:26:18.5 7.77(72) 81 G9.5IV 3.726(10) 0.34(8) 0.39(7) 0.6 0.32 30.6 -3.1 6 1.5 10 1.3 12 1.4 PMS
61
We can rule out most of the pre-MS candidates being Li-rich post-MS stars.
Based on the surveys of Li abundances in field subgiants by Randich et al. (1999)
and Pallavicini et al. (1987), we do not expect to find any post-MS subgiants with
EW(A6707) > 100 mA. Even if our measured EWs for the Li I A6707 line are
over-estimated due to low spectral resolution, the overestimate would have to
be greater than a factor of two to reconcile our sources with even the most Li-rich
subgiants found in the Randich et al. survey. Our spectral analysis suggests that
the majority of our sample stars are both Li-rich and above the main sequence
(i.e. pre-MS).
Could some of our stars be post-MS chromospherically active binaries (CABs)
or RS CVn systems? The light from an RS CVn system would be dominated by a
rapidly rotating, evolved (subgiant) primary. Only six of our targets are Li-poor
subgiants (HIP 63797, 81775, TYC 8293-92-1, 7833-1106-1, 7858-526-1, and 8285-
847-1). The first three appear to be normal subgiants. TYC 7833-1106-1 is possibly
a spectroscopic binary. TYC 7858-526-1 has a wide, broad Ho absorption line. It
appears to be a multiple late-F star (we classify it as F8.5; Houk (1982) classify it
as F5), so the star could hide a cosmic Li abundance due to the increased ioniza
tion of Li I in F stars (and correspondingly lower EW(Li)). The system could be
a legitimate member, but we exclude it from the pre-MS sample. The subgiant
TYC 8285-847-1 (HIP 69781 = V636 Cen) is probably a CAB. It is a previously
known grazing, eclipsing binary (e.g. Popper, 1966) and its saturated X-ray emis
sion argues for being a true CAB. Finally, the Li-rich star TYC 8992-605-1 (star
#19; KO+III) is the only RASS-ACT /TRC star that appears in the giant regime of
Fig. 2.1. The star is an obvious spectroscopic binary of nearly equal mass. We
believe this star is probably a pre-MS binary, and include it in our "pre-MS?"
Table 2.6—Continued
(1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11) (12) (13) (14) (15) (16) (17) (18) (19)
ID Name OB a, 5 (ICRS) T^SEC Pi Spec. log(Te/^-) logL/L© Av EW(HQ;) EW(Li) log(Lx) log(Lx/ DM97 PSOl SDFOO Class Notes
# TYC Grp h m s ° " (mas) % Type (K) (dex) (mag) (A) (A) (ergs 1) ^BOL) T, M r, M. T,M . . .
81 7852-51-1 UCL 16:18:38.56 -38:39:11.8 9.30(54) 5 F9IV 3.780(6) 0.33(5) 0.32(4) 2.6 0.13 30.3 -3.8 191.2 221.2 26 1.3 PMS
82 7857-648-1 UCL 16:21:12.19 -40:30:20.6 6.43(54) 1 G8IV 3.738(7) 0.23(8) 0.69(6) 1.4 0.25 30.5 -3.3 11 1.3 141.2 19 1.2 PMS i
83 7S57-514-1 UCL 16:23:29.55 -39:58:00.8 5.91(72) 91 G3IV 3.764(7) 0.11(11) 0.95(8) 2.0 0.20 30.4 -3.4 231.1 301.1 66 1.2 PMS? i
84 7853-227-1 UCL 16:27:30.55 -37:49:21.6 5.55(85) 66 G9.5IV 3.726(10) 0.21(14) 0.50(6) 0.6 0.35 30.5 -3.3 81.4 121.2 16 1.2 PMS
85 7353-2640-1 UCL 16:31:42.03 -35:05:17.2 7.01(69) 92 G7,5IV 3.742(9) 0.05(9) 0.33(5) 1.2 0.24 30.4 -3.3 19 1,1 25 1.0 28 1.1 PMS
86 7349-2191-1 UCL 16:35:35.99 -33:26:34.7 5.22(69) 7 K2-IV 3.687(7) 0.35(12) 0.34(6) 0.1 0.35 30.9 -3.0 21,3 41.5 5 1.6 PMS
87 7858-830-1 UCL 16:39:59.30 -39:24:59.2 4.63(75) 100 G4IV 3.760(10) 0.74(14) 0.27(7) 1.5 0.24 30,7 -3.3 61.8 51,7 9 1.6 PMS
88 7871-1282-1 UCL 16:42:24.00 -40:03:29.7 5.02(54) 92 GlIV 3.772(11) 0.92(10) 0.74(8) 0.3 0.18 30.7 -3.7 51.9 41.9 71.8 PMS
Note. — Columns: (1) short ID, (2) Tycho-2 name, (3) OB subgroup, (4) Tycho-2 position (ICRS, epoch 2000.0), (5) Secular parallax estimate (mas) (see §2.6.3),
(6) Membership probability (with = 1 kms~^), (7) Spectral type (see §2,4.1), (8) Effective temperature (see §2.6.2), (9) Luminosity (see §2.6.5), (10) Extinction
Av (see §2.6.5), (11) Equivalent Width of Ha A6562.8 (A) (see §2.4.2.1), (12) Corrected EW of Li I A6707.8 (A) (see §2.4.2.2), (13) X-ray luminosity (ergs-i) (see
§2.2.2), (14) Logarithm of ratio between X-ray and bolometric luminosities, (15) Age r (in Myr) and Mass MimM/MQ ) using DM97 tracks, (16) Age r (in Myr) and
Mass M. (in M/MQ) using PSOl tracks, (17) Age T (in Myr) and Mass M (in M/MQ) using SDFOO tracks, (18) Class - PMS = pre-main sequence, PMS? = pre-main
sequence? (see §2.5.1), (19) Notes: (a) Star #10 = HIP 59721 is a common proper motion pair (p = 24") wiUi HD 106444 = HIP 59716 (F5V, TT = 9.92 ± 2.53). (b) Star #19
= Einstein Slew Sircvey source lES 1233-63.4. (c) Star #23 = HD 311894 = 2E 1241.6-6315 = Eclipsing binary discovered in The All Sky Automated Survey (Pojmanski,
1998): ASAS J124435-6331.8 (Per. = 2.6 days), (d) Star #27 = 2E 1255.1-7012 = EUVE J1258-70.4. (e) Star #34 = PDS 66 = Hen 3-892 = IRAS 13185-6922. This is the lone
classical T Tauri star identified in this survey. [O I] A6300 is seen in emission (EW = 0.2A). (f) Star #46 = CCDM J1437-4145AB. Unresolved binary (p = 0.4"). (g) Young
ROSAT star identified by Wichmaim et aL (1997b) (h) Yoimg ROSAT star identified by Krautter et al. (1997) (i) Star #66 = CCDM J15370-3127A = RX J1537.0-3136 =
2E 1533.9-3126?. Our spectrum is of the bright (V = 10) primary; the BC components (both V = 13) are 4" away, and off-sUt. (j) Star #69 = CCDM J15394-2710AB.
Unresolved binary (p = 0.4"). (k) Star #76 = CCDM J16038-4356AB. Uiu-esolved binary (p = 0.3"). (1) Star #79. Faint companion off-sUt (p ~ 4").
63
Table 2.7. RASS-ACT/TRC & Hipparcos Stars Rejected as Sco-Cen Members
(1) (2) (3) (4) (5) (6) (7) (8)
Name a, S (ICRS) Pi SpT EW(Ha) EW(Li) log(Lx/ Type
h m s ° " % (A) (A) iX'bol)
TYC 8222-105-1 11:35:03.76 -48:50:22.0 71 F8.5V 2.7 0.19 -3.4 ZAMS
TYC 8982-3046-1 12:04:14.42 -64:18:51.7 97 GIV 2.0 0.20 -3.1 ZAMS
TYC 8990-701-1 13:13:28.11 -60:00:44.6 9 F9V 2.2 0.11 -3.4 ZAMS
TYC 8285-847-1 14:16:57.91 -49:56:42.3 43 G2,5IV 2.2 0.00 -3.9 subgiant (CAB)
TYC 7833-1106-1 15:08:00.55 -43:36:24.9 4 G2IV 0.6 0.00 -3.5 active subgiant
TYC 8293-92-1 15:09:27.93 -46:50:57.2 28 Korv 0.7 0.05 -3.4 active subgiant
TYC 7318-593-1 15:31:21.93 -33:29:39.5 94 G9.5V 0.0 0.15 -3.3 active dwarf
TYC 7858-526-1 16:38:38.47 -39:33:03.5 98 F8.5IV 2.5 0.02 -4.0 active subgiant (CAB)
HIP 63797 13:04:30.96 -65:55:18.5 9 G3.5IV 2.4 0.00 <-4.5 subgiant
HIP 65423 13:24:35.12 -55:57:24.2 99 GOV 2.1 0.18 -3.6 ZAMS®-
MP 68726 14:04:07.12 -37:15:50.5 60 G0.5in 1.9 0.00 <-5.1 giant
HIP 72070 14:44:30.96 -39:59:20.6 62 GIV 2.6 0.16 <-4.2 ZAMS
HIP 74501 15:13:29.22 -55:43:54.6 3 Gi.sin 2,4 0.03 <-4.9 giant
HIP 77015 15:43:29.86 -38:57:38.6 61 G0.5V 2.8 0.06 <-4.1 dwarf
HIP 79610 16:14:43.02 -38:38:43.5 37 G0.5V 2.7 0.03 <-4.2 dwarf
HIP 81775 16:42:10.36 -31:30:15.0 14 Girv 2.8 0.04 <-4.1 subgiant
Note. — Columns; (1) Name from Tycho-2 or Hipparcos catalogs, (2) position (ICRS, epoch 2000.0),
(3) Membership probability (with v^isp = 1 kms~^), (4) Spectral type (see §2.4.1), (5) Equivalent width
of HQ A5562.8 (A), (5) Corrected EW of Li I A6707.8 (A), (7) Logarithm of ratio between X-ray and
bolometric Ixuninosities (approximate upper limits assume RASS PSPC detection limit of 0.05 cts"^
and HRl = 0), (8) Class of object
^HIP 65423 = HD116402. Cutispoto et al. (2002) measure EW(Li) = 220 mA, and usini = 35 kms~^.
64
sample. It appears that CABs are a negligible contaminant when using X-ray and
kinematic selection in tandem with medium dispersion spectroscopy to identify
pre-MS populations.
2.5.3 Sample Completeness
We can make a rough estimate how many stars our selection procedure should
have detected by counting the number of massive Sco-Cen members in a certain
mass range, and assuming an initial mass function. We assume a complete mem
bership census within a limited mass range (the revised B-star Hipparcos mem
bership from Z99), and then extrapolate how many stars we should have seen
in our survey. We produce a theoretical H-R diagram for the subgroups' B stars
(discussed at length in §2.7.2), and calculate masses from the evolutionary tracks
of Bertelli et al. (1994) (Z = 0.02). We choose 2.5 MQ as our lower mass bound
ary (roughly the lower limit for B stars), and adopt 13 MQ as the upper mass
boundary (slightly higher than the highest inferred mass from the main sequence
members). In this mass range, we count 32 LCC members and 56 UCL mem
bers. We use a Kroupa (2001) IMF to predict how many low-mass stars might
belong to the OB subgroups. Down to the hydrogen-burning limit (0.08 MQ), a
total population of 1200i3oo stars in LCC and 2200 ± 300 stars in UCL is predicted
(Poisson errors)'*. Between 1.1-1.4 MQ, the mass range of a 15 Myr-old population
that our survey can probe (see §2.7.1 and Fig. 2.8), the Kroupa IMF predicts a
population of 291® stars in LCC, and Slly stars in UCL. In this mass range, our
survey detects 36 pre-MS stars in LCC and 40 pre-MS stars in UCL. The number
of observed pre-MS stars with 1.1-1.4M© corresponds to +l.lcr and -1.6cr of the
•'For low-number statistical uncertainties, we use the l a values from Gehrels (1986) throughout.
65
predicted number, for LCC and UCL respectively. This suggests that our survey
is fairly complete for LCC, but we might be missing ~10 members with masses
of 1.1-1.4 Mr.) in the more distant UCL subgroup if the subgroup mass function
is consistent with the field star IMF. The missing members of the UCL OB sub
group could be X-ray faint (L^ < 10^°-^ ergs ') stars which we were capable of
detecting in the closer LCC subgroup. The IMF extrapolation does suggest that
we have likely found at least the majority of stars in this mass range in both OB
subgroups (if not a complete census for LCC) and that our samples are represen
tative of the total population.
2.6 The FI-R Diagram
In order to investigate the star-formation history of the LCC and UCL OB sub
groups, we convert our observational data (spectral types, photometry, distances)
into estimates of temperature and luminosity. We then use theoretical evolution
ary tracks to infer ages and masses for our stars.
2.6.1 Photometry
The primary sources of photometry for our sample of association member can
didates are the Tycho-2 catalog (Hog et al., 2000a,b) and 2MASS catalog (Cutri
et al., 2003). However, the Tycho and 2MASS bandpasses are non-standard, and
must be converted to standard photometric systems to enable comparison with
intrinsic colors of normal stars and the interstellar reddening vector. To convert
the Tycho photometry to the Johnson system, we fit low-order polynomials to the
data in Table 2 of Bessell (2000) (relations given in Appendix C). A caveat is that
Bessell's calibrations are for B-G dwarfs and K-M giants. The majority of our
66
stars appear to be pre-MS G-K stars, whose intrinsic colors should more closely
match those of dwarfs rather than giants. To convert the 2MASS JHKg data to
the system of Bessell & Brett (1988), we use the conversions of Carpenter (2001).
The original optical and near-lR photometry for our target stars is given in Tables
2.2 and 2.1.
2.6.2 Temperature Scale
To fix stellar properties as a function of spectral type, we adopt relations (i.e. in
trinsic colors, BCs) from Table A5 of Kenyon & Hartmann (1995). Previous stud
ies have shown that colors and BCs as a function of 11 f/ are largely independent
of surface gravity over the range of interest for this study (e.g. Bessell, Castelli,
& Plez, 1998). However, Te// decreases with lower log 5 for FGK stars. After
some investigation (see Appendix A), we decided to adopt the dwarf Tgff scale
of Schmidt-Kaler (1982) (which Kenyon & Hartmann (1995) also use) with a -35 K
offset to account for the effects of lower log g in our sample stars. The scatter
in published dwarf T, F J scales is 60 K (la) among G stars, so while the shift is
systematic, its magnitude is of the order of the uncertainties. The uncertainties in
given in column 9 of Tables 5 and 6 include the uncertainty in spectral type
and the scatter in published Tg// scales. The typical la uncertainties in 7^// for
the pre-MS stars is sal GO K.
2.6.3 Secular Parallaxes
All of the stars in our sample have published proper motions, but only a few
dozen have trigonometric parallaxes measured by Hipparcos. The stars are dis
tributed over hundreds of square degrees of sky, and inhabit stellar associations
67
which are tens of parsecs in depth. Adopting a standard distance for all of the
stars in the association introduces unwanted scatter in the H-R diagram. With
accurate proper motions available, we calculate individual distances to the pre-
MS candidates using moving cluster or "secular" parallaxes (e.g. Smart, 1968).
We adopt the equations and formalism of de Bruijne (1999b), as well as his space
motions and convergent points for the LCC and UCL OB subgroups. The uncer
tainties in the secular parallaxes are dominated by the uncertainties in the proper
motion (cr„ oc a^), but contain a term added in quadrature accounting for a pro
jected lkms~^ internal velocity dispersion (see §4 of de Bruijne, 1999b). The
secular parallax is only meaningful if the star is indeed a member of the group.
Our spectroscopic survey has confirmed that most of the candidate stars are le
gitimately pre-MS, and that they are most likely members of the OB subgroups.
Secular parallaxes for older, interloper stars are meaningless and ignored. In Ta
bles 2.5 and 2.6, we list the secular parallaxes and membership probabilities for
the pre-MS stars in our survey. We calculate membership probability Pi (using
fonnulae 4 and 6 from Z99), which have assumed an internal velocity dispersion
of 1 kms"^ (de Bruijne, 1999b; Madsen, Dravins, & Lindegren, 2002) respectively.
We independently confirm that 1 kms"^ is a good estimate of the internal veloc
ity dispersion for low-mass stars in §2.6.4.
The robustness of our method can be illustrated (Fig. 2.4) by comparing the
secular parallaxes {Ksec) to the Hipparcos trigonometric parallaxes (•khip)- The un
certainties are typically 1-2 mas for the Hipparcos parallaxes, and 0.5-1 mas for our
secular parallax estimates. The secular and trigonometric parallaxes agree quite
well for the few pre-MS stars in our sample for which Hipparcos measured the
parallax. The secular parallaxes yield distance uncertainties of ~5-15 % for most
68
of the pre-MS stars.
15
75 pc
10 - lOOpo
m <53
s u 0) CO
g _200pc
_ 250 pc
0 0 10
[mas]
Figure 2.4: Comparison between Hipparcos astrometric parallaxes and our secular
parallaxes calculated using the moving group method. Data points are Pre-MS
(and Pre-MS?) association members from Tables 2.6 and 2.5.
2.6.4 Independent Confirmation of Vdiap ~ 1 km s~'for Low-Mass Members.
With a new, high-quality astrometric catalog now available (Tycho-2), one can
address the question: is the internal velocity dispersion of the post-T Tauri mem
bers the same as that for the high mass members (~1 kms"^^)? We would like
to answer this question as a check. The secular parallaxes that we calculate in
§2.6.3 assume that the subgroup velocity dispersions for the low-mass stars are
the same as that observed for the high mass stars (~1 km s''; de Bruijne, 1999b;
69
Mad sen, Dravins, & Lindegren, 2002). In principle, one can calculate an upper
limit to the velocity dispersion with the Tycho-2 astrometry alone.
As previously demonstrated, pre-main sequence members of the OB sub
groups can be efficiently selected by their strong X-ray emission and convergent
proper motions. To identify low-mass member candidates, we construct a cross-
referenced catalog of all Tycho-2 stars with ROSAT All-sky Survey BSC and FSC
X-ray sources within 40" radius (hereafter RASS-TYC2), and analyze the distri
bution of their proper motions. Within the LCC and UCL regions (defined by de
Zeeuw et al. 1999), we find 271 RASS-TYC2 stars with {13 - V) > 0.60 (G-type or
later) in LCC and 328 in UCL. For simplicity, we do not apply a magnitude re
striction other than the Tycho-2 magnitude limit. The vast majority are consistent
with being pre-MS or ZAMS at d = 100-200 pc.^
To search for subgroup members, we plot the proper motions for the RASS-
TYC2 stars in (/i„, Hr) space instead of (/i„, ns). The proper motion components
represent the motion toward the subgroup convergent point (/i„) and perpendic
ular to the great circle between the star and the convergent point (/u-r) (Smart,
1968). The expectation value of Hr for an ensemble of bona fide cluster members
is zero, and jiy scales with distance and angular separation from the convergent
point. We adopt the space motion for the LCC and UCL subgroups from de Brui-
jne (1999b) (the gu^ = 9 solutions in their Table 5, where the space motion vec
tor can be converted to a convergent point solution using eqn. 10 of de Bruijne
^In principle, this analysis could have produced an original list of pre-MS candidate subgroup members independent of Hoogervverf's (1999) work. The availability of improved astrometry (namely the Tycho-2 catalog; made publicly available early 2000), and a deeper catalog of X-ray sources (the Rosat All-Sky Faint Source Catalog; made publicly available mid-2000), have enabled this analysis.
70
(1999b)). Low-mass members of the OB subgroups stand out dearly in (/i„, jir)
space (Fig. 2.5). The mean subgroup proper motion values (de Bruijne 1999a; Ta
ble 4) are shown as dashed vertical lines, where p. for members is approximately-
equal to We define boxes around the loci in Fig. 2.5 to select probable associ
ation members for statistical study (70 stars in LCC, 105 in UCL).
40
20
LCC
U >^
0 VI 0 (0
20
40 - 2 0 20 40
pt„(masyr )
60
UCL
'• •« • >»• •«?ii v
20 40
u (mas yr )
Figure 2.5: Proper motions of RASS-TYC2 stars with (B — V) > 0.60, lying within
the boundaries of the LCC and UCL OB subgroups defined by de Zeeuw et al.
(1999). The mean /2„ values for the OB subgroups are shown by vertical dashed
line (from de Bruijne 2000). Association members should have a mean value of
fir = 0, with small scatter a{nr).
In order to estimate the observed dispersion in in a way that is insensi
tive to the boundaries of the subjectively drawn selection box (Fig. 2.5) and the
71
10 LCC (7(/i^) = 2.89 ± 0.46 mas yr~'
5 i:
0
-5
- 1 0 - 2 1 0 1 2
Standard Deviation
: UCL - ~ 3.48 ± 0.49 mas yr"
-5
- 2 2 - 1 0 1
Standard Deviation
Figure 2.6: Probability plots for for stars in the boxes defined in Fig. 2.5. A
Gaussian profile will produce a straight line, where the slope gives the standard
deviation. The observed distributions are consistent with the Tycho-2 proper mo
tion errors combined with intrinsic velocity dispersions of 0.6^[J;e km s" (LCC)
and 1.0i?;g kms ' (UCL).
presense of outliers, we use probability plots (Fig. 2.6). We follow the analysis
method outlined in §3 of Lutz & Upgren (1980). In a probability (or "probit") plot,
the abscissa is the expected deviation from the mean predicted for the ith sorted
data point in units of the standard deviation, and the ordinate is the data value
in question (nr)- The slope of the probability plot distribution yields the standard
deviation, and the y-intercept is the median. To fit a line to the probability plots
in Fig. 2.6, we use the Numerical Recipes least-squares routine fit, and trim 10%
from both sides of the distribution to mitigate against the effects of outliers. The
72
probability plots yield standard deviations of cr(/i^) = 2.9 ± 0.5 masyr " ' (LCC)
and 3.5 ±0.5 masyr"' (UCL). The mean values of Hr are close to zero (0.7 ± 0.3
masyr"^ for LCC, 0.3 ± 0.3 masyr for UCL), consistent with the expectation
that most of the RASS-TYC2 stars in the boxes in Fig. 2.5 are subgroup members.
An upper limit to the velocity dispersions of the low-mass subgroup member
ships can be estimated as follows. The observational errors in range widely
from 1-7 masyr~^ (3.0 ±0.9 masyr"^). We use Monte Carlo simulations to esti
mate the observed scatter expected in accounting for both the Tycho-2 proper
motion errors and the intrinsic velocity dispersion of the association. The intrin
sic velocity dispersion (in kms"' ) translates into a proper motion dispersion
cr(//^"^) (in mas yr~') as a function of mean subgroup parallax tt (adapted from de
Bruijne (1999) eqn. 20):
= TTtJA (2 .1 )
where A = 4.74 kms"' yr~^, the astronomical unit expressed in km divided by
the seconds in a Julian year. For the simulations, we model velocity dispersions
ranging from cr*„j = 0-3 kms"' in 0.5 kms"' steps. We adopt the mean distances
to LCC and UCL from de Zeeuw et al. (1999). We generate 10^ Gaussian deviates
for each star with zero mean and a standard deviation equal to the square root
of the observed value of cr(/x^) and the model (7(/x"'') values added in quadrature.
Statistical testing showed that clipping the Monte Carlo data at the box bound
aries in Fig. 2.5 ([JUT-I < 8 masyr"^) had negligible effect on the probability plot
determinations of a(/ir), so all Monte Carlo values were retained.
73
Table 2.8. Estimates of cr{(ir) from Monte Carlo Simulations and Observations
(1) (2) (3)
Sample LCC UCL cril^ r ) (masyr~i) (masyr"' )
Observed RASS-TYC2 sample 2.89 ±0.46 3.48 ±0.49
Model = 0.0 kms~i 2.62 3.18
Model cr*^^ = 0.5 kms~^ 2.78 3.27
Model = 1.0 kms~^ 3.24 3.48
Model = 1.5 kms~^ 3.84 3.89
Model a* . = 2.0 knis"' znt 4.53 4.35
Model (T* , = 2.5 kms~^ int 5.30 4.86
Model = 3.0 kms~^ 6.08 5.41
A comparison between the a (fir) values for the Monte Carlo simulations and
the observations is shown in Table 2.8. It appears that the internal velocity dis
persions of the subgroups are indeed detectable. The observed oiiir) values for
LCC and UCL are consistent with internal velocity dispersions of = 0.6l'o;g
kms~' and 1.0ti;o kms"S respectively. The 95% confidence level upper limits
to the velocity dispersions are <1.6kms~' (LCC) and <2.2kms' ' (UCL). We can
rule out velocity dispersions of 3kms"*^ (de Zeeuw et al. 1999), as this would
have produced a dispersion of cr(/x,-) ~ 5-6 mas yr'"^ in both subgroups.
The velocity dispersions determined from the Monte Carlo simulations are
strictly upper limits only The MML02 survey observed most of the stars in the se
lection boxes in Fig. 2.5, and some of those are known not to be pre-MS members.
Interlopers will evenly populate the Fig. 2.5 selection boxes, leading to slightly
inflated dispersions in although the use of probability plots largely mitigates
against this effect. Taking into account the lack of spectroscopic confirmation of
74
pre-MS status for all of the RASS-TYC2 candidate members. We conservatively
conclude the following: The intrinsic velocity dispersion ofpost-T Tauri stars in
the LCC and UCL OB subgroups i s <2 kms" \ and cons i s t en t w i th t he va lue (1 kms~^ )
measured for the early-type members by de Bruijne (1999b) and Madsen, Dravins, &
Lindegren (2002). Currently, there is no evidence for mass-dependent dynamical
evolution. A future radial velocity survey of the subgroup members will enable
us to determine whether these unbound associations are expanding, and possibly
determine an independent "expansion age" estimate.
2.6.5 Luminosities
With five-band photometry, a temperature / spectral type estimate, and a sec
ular parallax, we calculate stellar luminosities for the pre-MS candidates. We
adopt the absolute bolometric magnitude of the Sun (Mboi© = 4.64) from Schmidt-
Kaler (1982). In order to compromise between the uncertainties in luminosity due
to reddening, photometric uncertainties, and possible K-band excess, we calcu
late the Mboi using the dereddened 2MASS J magnitude. We estimate the vi
sual extinction from a weighted mean of Ay estimates from the color excess in
(B — y)and {V — J). We took the E(B — V) formula from Drilling & Landolt
(2000) and the value of Aj/Av{= 0.294) was taken from the near-IR extinction
law of Mathis (1990) for a central wavelength 1.22 //m. The reddening Aj typ
ically ranged from 0 to 0.35 mag with formal uncertainties of ~0.1 mag. The
typical uncertainty in logL/Lc, for the pre-MS candidates is ?«0.08 dex. With the
luminosities and X-ray fluxes from the RASS BSC catalog (Voges et al., 1999), we
calculate the ratio of X-ray/bolometric radiation for the stars with X-ray coun
terparts. The derived values of log{Lx /l-boi) are in the range of 10"^-^ - lO"^-®,
indicating coronal X-ray emission elevated above most ZAMS G-type stars (e.g.
75
Pleiads; Stauffer et al., 1994).
2.6.6 Evolutionary Tracks
In order to infer theoretical masses and ages from our pre-MS candidates, we use
the evolutionary tracks from D'Antona & Mazzitelli (1997) (DM97; Z = 0.02, Xo -
2x10"''), Siess, Dufour, & Forestini (2000) (SDFOO; Z = 0.02), and Palla & Stabler
(2001,; PSOl). Ages and masses for a given log Tgj/ and logL/Lg were calculated
using an interpolation algorithm. Given the mean observational errors (cr(logTe//,
l o g L / L © ) = 0 . 0 0 7 , 0 . 0 7 8 d e x f o r L C C P r e - M S s t a r s , a n d a { l o g T e f f , l o g L / L - v ) =
0.009, 0.084 dex among UCL Pre-MS stars), we estimate the isochronal age un
certainties for an individual star to be approximately 4, 5, 7 Myr (DM97, PSOl,
SDFOO) in LCC, and 4, 5, 5 Myr (DM97, PSOl, SDFOO) in UCL, as illustrated in
Fig. 2.7. The uncertainties in the interpolated masses are 0.1 MQ for all three sets
of tracks. Fig. 2.8 shows the H-R diagram for the pre-MS candidates overlayed
with the evolutionary tracks of DM97.
2.7 Results
The ages of the low-mass population of LCC and UCL have not been estimated
before, though de Geus et al. (1989) and de Zeeuw & Brand (1985) give main
sequence turn-off ages. In §2.7.1 we estimate the pre-MS ages for the two sub
groups, and put an upper limit on the intrinsic age spread. In §2.7.2 we calculate
new turn-off ages for the subgroups using early B stars from the revised Hipparcos
membership lists of Z99.
76
l.SxlO"
10"
5000
0 0 10 20 30 40
Age [Myr]
Figure 2.7; A histogram of the inferred ages from the DM97 and SDFOO tracks for
a hypothetical LCC Pre-MS star with average H-R diagram point (7e/;,logL/L
and gaussian uncertainties. The extreme right bin retains all points older than 40
Myr. The standard deviations are calculated using only stars with ages between
1-100 Myr.
2.7.1 Pre-MS Ages and Age Spread
The H-R diagram for our "pre-MS" and "pre-MS?" stars is shown in Fig. 2.8,
overlayed with the evolutionary tracks of D'Antona & Mazzitelli (1997). The tem
peratures and luminosities of the pre-MS stars are given in Columns 9 and 10 of
Tables 2.6 and 2.5, along with their inferred masses and ages (columns 15 through
17). One notices immediately that the bulk of isochronal ages are in the range of
10-20 Myr. The age range is nearly identical for both groups. To assess the ef
fects of our magnitude limit in biasing our mean age estimates, in Fig. 2.9 we
Average LCC HRD point log(Trf,) = 3.763±0.007, log(L/Lg) = 0,250±0.078
N = 100000 samplings
1 Age(DM97) =
17.6 ± 3.6 Myr
;e(SDF00) =
Myr 23.9 ±
77
GO G5 KG K2
1
60 O 1—^
• • 0
approx. Tycho limit
_® LCC Pre-MS
_ o UCL Pre-MS -0,5
3.75 3.7
log(T.„)
Figure 2.8: Theoretical H-R diagram for stars identified as Pre-MS or Pre-MS? in
Tables 2.5 and 2.6 in the UCL (open circles) and LCC samples (filled circles). The
pre-MS evolutionary tracks of DM97 are overlayed. The ACT/TRC magnitude
limit (V = 11 mag) is shown for a distance of 150 pc (Ay = 0.3 assumed). The
star in the bottom right comer (TYC 8648-446-1) is one of the faintest stars in our
sample (V = 11.2 mag) with larger than average errors in logL/Lv, - hence its
unusual position. The average la error bars in log Tg// and log L/L© are shown.
plot the mean pre-MS age (with standard errors of the mean) for the pre-MS sub
group samples as a function of minimum log Tg// cut-off. The magnitude bias of
our survey is clearly apparent: the mean age systematically decreases when stars
with logTe// < 3.73 are included in the calculation. In calculating the pre-MS
ages of the OB subgroups, we explicitly omit the pre-MS stars with log T^,;/ <
3.73 (30% of our sample). This temperature threshold intersects our magnitude
78
limits at ages of ~25 Myr for stars of 1M© on the DM97 tracks. That the lines
in Fig. 2.9 are nearly flat for log T^// > 3.73, suggests that (detectable) stars with
ages of >25 Myr are not a significant component of either subgroup (also see dis
cussion in §2.8.1).
Fig. 2.10 displays histograms of the isochronal ages for the pre-MS stars in the
LCC and UCL subgroups derived using DM97 and SDFOO evolutionary tracks.
These tracks represent the extrema in age estimates for our sample (DM97 is
youngest, SDFOO is oldest). The la age dispersion among the unbiased samples
(log Tf.f f > 3.73) is 5-9 Myr for both groups. If we remove the known spectro
scopic binaries (see notes in Tables 2.6 and 2.5), the age dispersions are 4-8 Myr.
Because there may be additional unresolved binaries, this observed age spread
places an upper limit on the intrinsic age spread. As illustrated in Fig. 2.7,
the individual H-R diagram positions of the pre-MS samples have logTe// and
log L/L© errors which fold onto the evolutionary tracks with age uncertainties
of 4-7 Myr. From this analysis, we conclude that the intrinsic la age dispersions
in each subgroup must be less than 2-8 Myr (i.e. ~2/3rds of the star-formation
took place in <4-16 Myr). Using the DM97 pre-MS ages, which agree best with
the turn-off age estimates (§2.7.2), we find intrinsic la age dispersions of 2 Myr
(LCC) and 3 Myr (UCL).This implies that 68% of the loiv-mass star-formation took
place within <4-6 Myr, and 95% within <8-12 Myr in the OB subgroups. Our ob
servational uncertainties and lack of knowledge about the unseen binarity of the
pre-MS sample stars do not allow us to constrain the age spread more precisely
than this. The mean age estimates for our unbiased pre-MS samples (logTe//>
3.73, SBs removed) are shown in Table 2.9. Counter to previous studies, we find
that LCC is slightly older than UCL by 1-2 Myr (at 1-3CT significance), indepen-
79
18
^ 16
s k—J (U Ofl < c to 14 cu
12
Figure 2.9: Illustration of the effects of magnitude bias on our mean age estimates
for the pre-MS populations. The abscissa is the minimum log(Teff) threshold for
evaluating the mean sample ages (using DM97 tracks). The ordinate is calculated
mean age with standard errors of the mean (shown; typically Myr). At cooler
temperatures (later than KO), the magnitude limit of our survey biases the sample
towards more luminous stars, thereby decreasing the mean age estimate. From
this diagram, we choose log T,,// = 3.73 (vertical dashed line) as the lower T,,//
cut-off for evaluating the mean pre-MS ages. Known spectroscopic binaries are
included here, but excluded in the final age estimates presented in Table 8. The
observed isochronal ages and spread are 16 ±5 Myr for LCC and 14 ±5 Myr for
UCL.
• LCC
OUCL
3.74 3.72 3.7 3.68 3.76
minimum log(T „) for pre-MS sample
80
30
20
10
0
r I I I ; I I I I I I I I I j I I I I
h LCC DM97 Tracks
- <T> = 17.2±0.9 (s.e.> ±4.2 (Ict)
0 10 20 30 40
Age [Myr]
30 I I I 1 I I I 1 I 1 I I I I I I I I I I - UCL
DM97 Tracks I - <T> = 14.7±0.8 ( s . e . y
20 h ±4.9 (la)
10 20 30
Age [Myr]
30
20
10
0
I I I I I I I I i I I I I I 1 I
LCC SDFOO Tracks <T> = 22.5±1,7 (s.e.)-
±7.8 (la)
0 10 20 30 40
Age [Myr]
30
20
10
I I I I I I t I j I 1 1 I I I I I f- UCL
SDFOO Tracks - <T> = 21.7±1.3 (s.e.>
±7.6(la)
10 20 30
Age [Myr]
40
Figure 2.10: Histograms of the isochronal ages for pre-MS and "pre-MS?" can
didates in Tables 2.5 and 2.6 from the models of DM97 and SDFOO. The filled
bins are for stars with logTg// > 3.73, and the unfilled bins are for the entire
(magnitude-biased) sample. Mean isochronal ages (with standard errors of the
means and la uncertainties) are given for the unbiased sample (log T^jf > 3.73).
Outliers with isochronal ages of >40 Myr are counted within the 40 Myr bin.
dent of which evolutionary tracks we use. From Fig. 2.10, we also conclude that
star-formation ceased approximately ~5-10 Myr ago in the subgroups.
2.7.2 Turn-off Ages
De Geus et al. (1989) published the most recent age estimates for the LCC and
UCL groups, but in light of the new Hvpparcos distances and subgroup member-
81
Table 2.9. Age Estimates of LCC & UCL
Ref. Tracks method age(LCC) age(UCL)
(Myr) (Myr)
1 4 pre-MS 17±1 15±1
1 5 pre-MS 21±2 19±1
1 6 pre-MS 23 ±2 22±1
1 7 turn-off 16 ±1 17±1
2 8 turn-off 11-12 14-15
3 8,9 turn-off 10-11 12-13
Note. — Uncertainties are standard errors of the
mean. Pre-MS age estimates exclude known SBs and
stars with log T,,/j < 3.73, which bias the calculated
ages. The turn-off ages from this work are deter
mined using only early-B stars classified as members
by de Zeeuw et al. (1999).
References. — (1) this work, (2) de Geus et al.
(1989), (3) de Zeeuw & Brand (1985), (4) D'Antona &
Mazzitelli (1997), (5) Palla & Stabler (2001), (6) Siess,
Dufour, & Forestini (2000), (7) Bertelli et al. (1994), (8)
Maeder (1981), (9) Cogan (priv. comm.)
82
ship lists, we feel it is worthwhile reevaluating the subgroups' turn-off ages. We
construct a theoretical H-R diagram for the B-type subgroup members of UCL
and LCC listed both in Table CI of Z99 and Tables A2 and A3 of de Bruijne
(1999b). Several of the "classical" ^ members rejected as members by Hipparcos
from Z99 are included as well. For input data, we use the following databases in
order of availability: (1) ubvyP photometry from the database of Hauck & Mer-
milliod (1997), (2) UBV photometry from Slawson, Hill, & Landstreet (1992), and
(3) UBV photometry from SIMBAD. For distances we use the secular parallaxes
{•Ksec) given in column 4 of Tables A2 and A3 of de Bruijne (1999b) when available,
or the Hipparcos parallaxes {TTHIP)- We deredden the stars with ubvy/S photometry
to the B-star sequence of Crawford (1978) using the prescription of Shobbrook
(1983). For stars with Stromgren photometry, we calculate Tg f/ using the tem
perature relation of Napiwotzki, Schonbemer, & Wenske (1993). If no ubvyP pho
tometry was available, we use UBV photometry to calculate the reddening-free
index Q (Crawford & Mandwewala , 1976) to infer the star's unreddened color.
A polynomial fit to Table 15.7 from Drilling & Landolt (2000) is used to calculate
Teff as a function of {B - V)o- The BC versus T^ff relation of Balona (1994) is
used for all stars. We linearly interpolate between the isochrones from Bertelli et
al. (1994) (Y = 0.28, Z = 0.02, convective overshoot) to infer ages for the subgroup
B stars. The theoretical H-R diagram is shown in Fig. 2.11.
UCL has a well-defined MS turn-off composed of the Hipparcos members HIP
67464 {u Cen; B2IV), HIP 68245 {(f) Cen; B2IV), HIP 68282 Cen; B2IV-V), HIP
71860 {a Lup; B1.51II), HIP 75141 {6 Lup; B1.5IV), HIP 78384 {rj Lup; B2.5IV),
and HIP 82545 (/i^ Sco; B2IV), as well as "classical" members (but Hipparcos non-
'"'Classical" members are early-type stars which were included in Sco-Cen membership lists before the Hipparcos studies of Z99.
83
5
4
® J \ J M
3
2
4.5 4.4 4.3 4.2 4.1 4
log(T,„)
Figure 2.11: Theoretical H-R diagram for the B-star candidate members of the
LCC & UCL memberships using the evolutionary tracks of Bertelli et al. (1994).
Only the most massive Hipparcos members were included in the age estimates.
The unusual variable HIP 67472 (n Cen; B2Vnpe; log L/L,.„ logTe// = 4.43, 4.0)
was excluded from the UCL turn-off age estimate.
members) HIP 82514 (fii Sco; Bl.SVp) and 73273 {j3 Lup; B2III)^. The variable star
HIP 67472 (ji Cen; B2Vnpe) was excluded. Using the Bertelli et al. (1994) tracks,
the mean age of the 7 turn-off Hipparcos members is 17 ± 1 Myr. Including the
two "classical" members has negligible effect on the mean age estimate. Our MS
turn-off age estimate for UCL is slightly older than de Geus et al.'s (14-15 Myr),
and is close to the mean pre-MS ages that we found in §2.7.1 (15-22 Myr).
''Note that the two Hipparcos non-members were found to be probable members by Hoogerw-erf (2000) if the long-baseline ACT (HIP 73273) and TRC (HIP 82514) proper motions were used instead of the Hipparcos values.
log(age) = 6.9 Myr
10 Myr 7.0
IS.Myr
16jyiyr 7.2
20 Myr 7.3
25 Myr 7.4
Myr 7.5
o«
- • LCC (HIP members)
^ LCC classical (HIP non-mem.)
O UCL (HIP members)
" UCL classical (HIP non-mem.)
Oo
84
LCC lacks a well-defined tum-off, however we have enough early-B stars in
the middle of their MS phase with which to make an age estimate. We estimate
the age for LCC from the following main sequence B stars: HIP 59747 {6 Cru;
B2IV), HIP 60823 (a Cen; B2V), HIP 61585 (o Mus; B2IV-V), HIP 63003 (//i Cru;
B2IV-V), and HIP 64004 Cen; B1.5V). The mean age for these five stars is 16 ± 1
Myr; similar to what we found for UCL, and it agrees well with the younger
end of our pre-MS age estimates (17-23 Myr). This age estimate is significantly
older than previous estimates (10-12 Myr), and warrants more critical examina
tion (§2.8.3).
The new results yielded by our age analysis of the OB subgroups are: (1) two-
thirds of the low-mass star-formation in each subgroup took place in less than a
~5 Myr span (and 95% took place within ~10 Myr), (2) the pre-MS and B-star
ages for LCC and UCL are in approximate agreement, (3) the B-star subgroup
memberships defined by Hipparcos have ages of 16 ± 1 Myr and 17 ± 1 Myr for
LCC and UCL, respectively. We discuss the implications of these results in §2.8.
2.7.3 The Census of Accretion Disks
An important question both for star and planet formation is the lifetime of ac
cretion disks around young stars. Statistics for the frequency of active accretion
disks around low-mass stars come predominantly from near-IR surveys of young
associations and clusters (Hillenbrand & Meyer, 1999; Haisch, Lada, & Lada,
2001a). The samples of low-mass stars surveyed are dominated by embedded
associations with ages of <3 Myr (e.g. Tau-Aur, Cha I, etc.), and older open clus
ters with ages of 30-100 Myr (e.g. Pleiades, IC 2602, a Per, etc.). Few well-studied
85
pre-MS stars of 3-30 Myr-old ages have been surveyed. The situation has recently
been slightly ameliorated by the discoveries of the TW Hya association and r]
Cha cluster (Kastner et al., 1997; Webb et al., 1999; Mamajek, Lawson, & Feigel-
son, 1999). Yet these samples are small (-^10-20 stars) and dominated by K and
M-type stars with masses of 0.1-0.8 MQ. Our pre-MS star sample is unique in its
mass (~1-1.5 A/Q) and age range (~ 10-20 Myr), so measuring its disk-frequency
provides a valuable datum.
Stars with EW(Ha) > lOA in emission are usually called Classical T Tauri stars
(CTTSs), which show spectroscopic signatures of accretion as well as near-IR ex
cesses (e.g. Hartigan, Edwards, & Ghandour, 1995). Stars lacking the strong Ha
emission and near-IR excesses are called Weak-lined T Tauri stars (WTTSs). This
can be explained as a correlation between the presence of magnetospheric accre
tion columns and an inner accretion disk (e.g. Meyer, Calvet, & Hillenbrand, 1997;
Muzerolle, Hartmann, & Calvet, 1998). Our H« EW measurements are discussed
in §2.4.2.1, and here we quantify the K-band excess of our targets. We calculate
the intrinsic {J - KG) color excess E(./ - KG)^ as defined by Meyer, Calvet, & Hil
lenbrand (1997):
E { J - A')o = ( J - K ) , b s - { J - K ) o - A y X { A J - A k ) (2.2)
where {J - Kg)obs is the observed color and (J - A'Jo is the intrinsic color of an
unreddened dwarf star of appropriate spectral type (Kenyon & Hartmann, 1995).
Uncertainties in each quantity were propagated in order to estimate the signal-to-
noise of the intrinsic color excess. The distribution of measured E(J-A'J,- values
indicate a systematic offset of a few hundredths of a mag. We subtract the small
86
offset, with the result being that the distribution of E(J-A's)o values is symmetric
about zero, with a few positive and negative ~2cr points. There is only one star
with a E( J — Ks)o color excess with S/N > 2.5: star #34 = TYC 9246-971-1 has an
intrinsic color excess of E(J — Ks)o = 0.26 ±0.06 implying a K-band excess. This
star also happens to be the only CTT identified in our optical spectra (EW(Ha)
= -39 A). TYC 9246-971-1 (= PDS 66, Hen 3-892) was originally identified as an
emission line star by Henize (1976), and classified as a CTT in the Pico dos Dias
survey of stars in the IRAS PSC catalog (Gregorio-Hetem et al., 1992). By virtue of
its position, proper motion, and spectral characteristics, we find that TYC 9246-
971-1 is a «8 Myr-old, ^1.2 MQ (DM97 tracks) member of the LCC subgroup.
Our secular parallax for TYC 9246-971-1 yields a distance of 861? pc; the third
nearest of the LCC pre-MS stars in our sample, and among the nearest CTTSs
known. Only 1/58 (1.7^'};°%; la Poisson) of pre-MS stars in LCC are classified as
bona fide CTTS, along with none (0/42) of the pre-MS members of UCL. For our
accretion disk frequency statistics, we use the isochronal ages derived from the
DM97 evolutionary tracks (which agree well with the turn-off ages), and include
the entire sample of 110 pre-MS stars (including the cooler stars which bias the
mean to younger ages). Only 1/110 (0.9lo;8°/o/ Icr) of 1.3 ± 0.2 (la) MQ stars with
ages of 13 ± 1 (s.e.) ± 6 {la) Myr are CTTSs. This implies that accretion terminates
in solar-type stars within the first 15 Myr of their evolution.
2.7.4 Is There a Sco-Cen OB Subgroup in Chamaeleon?
A recent study (Sartori, Lepine, & Dias, 2003, hereafter SLD03) proposed that an
other, previously undiscovered, and nearby, OB subgroup lay just to the south
of LCC in Chamaeleon. If the subgroup were real, the putative association could
possibly add a sigiiificant new piece to the puzzle of the star-formation history of
87
the Sco-Cen region.
SLD03 presented a membership list of 21 B stars in the Chamaeleon region
that they claim constitute a new OB subgroup of Sco-Cen (§2.3 and Table 5 of
their paper). The putative Cha OB members were selected solely by distance
(120-220 pc) and projected proximity to the Chamaeleon molecular clouds. We
present two observations which demonstrate that either the Cha subgroup mem
bership list of SLD03 is severely contaminated by field stars, or that the group
doesn't exist.
SLD03 measured a velocity dispersion for their Cha B-star sample of {au,
oy, aw) = (8, 11, 6) kms"'. Observations of nearby OB associations show that
their velocity dispersions are small - tjrpically <1 km s~^ (de Bruijne, 1999b; Mad-
sen, Dravins, & Lindegren, 2002). Torra, Fernandez, & Figueras (2000) find that
young (<100 Myr-old), nearby (d = 100-600 pc) field OB stars distributed all over
the sky have a velocity dispersion of {au, ay, aw) — (8, 9, 5) kms~'. This is simi
lar to the velocity dispersion for the Cha B stars, and suggests that if the sample
contains a bona fide OB association, it is probably severely contaminated by field
B stars.
There is also a discrepancy in the numbers of Cha OB members versus non-
members in the Chamaeleon region. If one searches the Hipparcos catalog for B
stars in the 180 deg' region surveyed for Chamaeleon ROSAT T Tauri stars by Al-
cala et al. (1995), constrained to distances between 120-220 pc, one finds that all 12
B stars within these constraints are considered Cha OB members by SLD03. How
many non-member field B-type stars would one expect? The projected density of
88
Hipparcos B stars with distances of 120-220 pc at Galactic latitude -18° is ~0.057
deg"'" (calculated in a 10°-wide band centered on b = -18°, covering all Galactic
longitudes). Over the 180 deg^ Cha region defined by Alcala et al. (1995), one ex
pects to find 10 ± VTd B-type field stars. One finds 12, consistent with the density
of field B-stars, and within the Poisson error bar. It is difficult to accept that all 12
B-type Hipparcos stars in this region are members of a new OB association, when
10 B-type field stars are predicted to exist. These numbers also suggest that there
is not a statistically significant over-density of B stars in the Chamaeleon region.
Along with the high velocity dispersion of the putative Cha OB membership list,
the evidence presented here suggests that there is no kinematic OB subgroup in
Chamaeleon, associated with Sco-Cen.
2.8 Discussion
We can address several interesting questions regarding the star-formation history
of Sco-Cen with data from our survey. Could the Sco-Cen progenitor giant molec
ular cloud (GMC) have produced a substantial population of low-mass stars for
an extended period (>5-10 Myr) before conditions were right to form an OB pop
ulation? Conversely, is there evidence for any low-mass star-formation after the
bulk of the high-mass OB stars formed? The OB star-formation in LCC and UCL
has apparently destroyed the progenitor GMC through supemovae and stellar
winds (e.g. de Geus, 1992; Preibisch & Zinnecker, 2000). However the region
is not totally devoid of molecular gas (e.g. the Lupus complex). We will first
examine whether there is any evidence of star-formation prior to the formation
of the OB subgroups (§2.8.1), and then assess the evidence for more recent star-
formation in the UCL region (§2.8.2). We will address the age of LCC in §2.8.3,
89
and discuss the formation of the subgroups in §2.8.4. Throughout the discussion,
we adopt the DM97 ages, as they agree more closely with the turn-off ages than
do the SDFOO and PSOl ages.
2.8.1 Is There Evidence for Star-Formation Before the Primary Bursts?
Is our survey sensitive to older stars which may have preceded the primary star-
formation episode? Three pre-MS stars in our sample have isochronal ages of >25
Myr (or undefined as lying below the ZAMS), however given the uncertainties
in Teff and logL/Lc, (Fig. 2.7), even a coeval «15 Myr-old population would
be expected to have statistical outliers. Here we explore three possible ways in
which older ZAMS stars could have escaped our attention.
One could argue that our surface gravity indicator is biasing our sample against
identifying ZAMS stars members (if they exist). If we disregard surface gravity
as a criterion, we gain only 4 more RASS-ACT/TRC stars (all between F8.5 and
Gl), and only one of those would have an isochronal age > 25 Myr (TYC 8222-105-
1; ~30 Myr). If they were legitimate, older members with real ages of > 25 Myr,
they should also be among the stars with the oldest isochronal ages in our sample,
which they are not. This suggests that their secular parallaxes, hence their lu
minosities, are unjustified, and that they are not members of the OB subgroups.
Coincidently, TYC 8222-105-1 is one of the earliest type stars in our sample (F8.5),
where our surface gravity indicator has the least fidelity (Fig. 2.1). We can state
that only one of the Li-rich stars showing dwarf gravity signatures that is co-
moving with LCC and UCL has an H-R diagram position and gravity suggestive
of ZAMS status.
90
If a significant ZAMS population existed in LCC and UCL, would our magni
tude and X-ray flux limits allowed their detection? X-ray surveys of the ZAMS-
age clusters IC 2602 and IC 2391 (—30-50 Myr) by Randich et al. (1995) and Patten
& Simon (1996) found that late-F and early-G stars ZAMS stars have X-ray lumi
nosities of Lx — 10^® ° - lO^"-® erg s"', with Lx/W ranging from 10"'^ ° to 10"^ ®.
The X-ray and optical flux limits imposed by the ROSAT All-Sky Survey and the
Tycho catalog allow us to detect ZAMS sources with Lx/Uo; > 10"^ '^ within 140
pc if they exist. If we adopt the X-ray luminosities of the G stars in IC 2602 and
IC 2391 as representative for a ~30 Myr-old population, we should have detected
roughly one-third of a putative Sco-Cen ZAMS population between masses of
1 and 1.2 MQ. We can put a rough upper limit on the number of >25 Myr-old
stars in our mass range. Assuming that TYC 8222-105-1 is a ZAMS member, and
that its H-R diagram position is not a statistical fluctuation from the locus of ~15
Myr-old stars, we detect one ZAMS star with age >25 Myr in the mass range
(1-1.2 Me). Accounting for the two-thirds of the ZAMS stars which would have
undetectable X-ray emission, and extrapolating over a Kroupa (2001) IMF, this
implies a population of ~100 stars with masses greater than 0.1 M©. This is <10%
of the stellar population predicted to exist in each OB subgroup (~1000-2000;
§2.5.3).
Could such ZAMS stars have left the region we probed? If we postulate that
the population was very centrally concentrated and gravitationally bound un
til the OB stars destroyed the giant molecular cloud (CMC) some ~10 Myr ago
(de Geus, 1992), then a 2 kms~ ' motion radially away from the subgroup cen
ter would have moved the star 20 pc in the past 10 Myr. This distance is the
approximate radius of both of the subgroups today (see Fig. 9 of Z99). Hence
91
if an older population was concentrated at center of the gravitationally-bound
GMC until the high mass stars destroyed the cloud, we would find them within
the projected boundaries of the subgroups so long as they inherited velocities of
<2 km s ^ The kinematic selection procedure of Hoogerwerf (2000) would have
selected such stars, since a large velocity dispersion (3 kms'"^) was initially as
sumed.
We conclude that there is no evidence for significant star-formation in the LCC
and UCL progenitor giant molecular clouds before the primary star-formation
episodes. Our findings are consistent with the idea that molecular clouds form
stars over a wide range of masses, and dissipate within timescales of ~10 Myr.
2.8.2 On-going Star-Formation?
Two obvious sources of young stars may be contaminating the UCL pre-MS sam
ple. The youngest, unembedded OB subgroup of Sco-Cen is Upper Sco (US),
with a nuclear age of 5-6 Myr (de Geus et al., 1989). US borders UCL near Galac
tic longitude 343°, and its space motion and distance are very similar to that of
UCL (Z99). The Lupus molecular clouds are also in the western region of UCL
(roughly between 335° < i < 345° and +5° <b < 25°). The T Tauri star population
within the major Lupus clouds was surveyed by Hughes et al. (1994), and the re
gion was recently mapped in ^CO by Tachihara et al. (2002). Dozens of pre-main
sequence stars were identified outside of the main cores by a pointed ROSAT sur
vey (Krautter et al., 1997), and the All-Sky Survey (Wichmann et al., 1997b). The
clouds lie at d = 140 pc (Hughes, Hartigan, & Clampitt, 1993), situated spatially
between the US and UCL subgroups of Sco-Cen (both d ~ 145 pc). Fig. 2.12 illus
trates the positions of the primary Lupus clouds, the pre-ROSAT Lupus T Tauri
92
star population, the pre-MS stars from our survey, and the B-star population of
the OB subgroups.
How does the presence of US and the Lupus molecular clouds (and their as
sociated T Tauri stars) affect our findings regarding the mean age of the UCL sub
group? We split our unbiased (log T^// > 3.73) "pre-MS" and "pre-MS?" members
of UCL into two groups using the Galactic longitude line 335° as a division. Most
of the molecular cloud mass in the Lupus region lies between 335° < i < 345°
(see Fig. 2 of Tachihara et al. (2002)). Using the DM97 tracks, we find that the
"eastern" UCL pre-MS sample surrounding the Lupus clouds has a mean age
of 13 ± 1 Myr, while the "western" UCL sample is somewhat older (16 ± 1 Myr).
The UCL stars with ages of <10 Myr are found in greater numbers near the Lupus
clouds and US border, supporting the idea that our UCL sample is probably con
taminated by more recent star-formation. The age estimate of UCL for the stars
west of i = 335° is probably more representative of the underlying UCL popula
tion.
Three of the youngest stars (HIP 81380, TYC 7858-830-1, and TYC 7871-1282-
1; 5-9 Myr; DM97 tracks) in our entire survey are positioned near a clump of 8
B stars at (i, b) = (343°, +4°). These three pre-MS stars also have secular parallax
distances of ~200pc, similar to what de Bruijne (1999b) found for the group of B
stars. The secular parallaxes may be biased, however, if this clump has slightly
different kinematics than the average UCL motion. This clump may represent
substructure within UCL. However de Bruijne (1999b) was unable to demonstrate
that the clump had distinct kinematics or age. The mean Hipparcos distance of the
clump B stars is 175 pc, with HIP 82514 (fj} Sco; B1.5Vp) and HIP 82545 {fj? Sco;
93
20
cio 10 CD
T3
rO
0
- 1 0
Figure 2.12: Map of the UCL and LCC subgroups of the Sco-Cen OB association
(Sco OB2). The B-star papulation from de Zeeuw et al. (1999) is shown by filled
squares. The pre-MS (and pre-MS?) sample from this survey is shown as open
circles. Fre-ROSAT T Tauri stars in the HBC catalog associated with the Lupus
cloud are shown as Xs (Herbig & Bell, 1988).
T r 1 I r 1 r
U C L N
_ us o
4 X
I v ° » " ; 0° ° • O •
O X oo
• • •o o • _
4, i
0 01
LCC
6^ (J
' J " o B, C • *
o
o Pre-MS stars (this study) • B stars x Lupus TTSs (HBC)
m Oo • •
1 .
J L J L J L
+
H
H
H
+
H
H
t l"
H
R R
340 320
1 ( d e g )
300
94
B2IV) as the most massive members. Our identification of 3 new pre-MS stars in
the same region with similar secular parallaxes supports the notion that this may
be a separate subgroup.
Some of our pre-MS stars were also identified in the ROSAT surveys of the
Lupus region by Krautter et al. (1997) and Wichmann et al. (1997b) (see notes in
Table 2.6). The presense of a significant population of pre-MS stars outside of star-
forming molecular clouds has been attributed by various authors to be due to one
or more of the following: (1) slow diffusion (1-2 kms "^) from existing molecular
clouds (e.g. Wichmann et al., 1997a), (2) ejection from small-N body interactions
(Sterzik & Durisen, 1995), (3) formation in situ from short-lived cloudlets (Feigel-
son, 1996), or (4) fossil star-formation associated with the Gould Belt (e.g. Guill-
out et al., 1998). Wichmann convincingly showed that most of the young RASS
stars in the Lupus region are at a distance of around ~150pc (similar to previ
ously published distances for the Lupus clouds and UCL), and that the stars are
roughly 10 Myr-old. Wichmann concludes that the dispersed pre-MS population
is most likely a manifestation of the Gould Belt. The OB subgroups of Sco-Cen are
major sub-structures of the Gould Belt (as defined by age and kinematics; Frogel
& Stothers, 1977), i.e. UCL is the dominant Gould Belt substructure in the Lu
pus region. We interpret the presence of dozens of pre-MS stars near the Lupus
clouds to be primarily the low-mass membership of the UCL OB subgroup. Our
analysis suggests that younger US or Lupus stars are a minor contaminant to our
UCL sample.
95
2.8.3 Is LCC Older than UCL?
Although our pre-MS and turn-off age estimates for LCC agree rather well, they
are substantially older (by ~50%) than previous values. The de Geus et al. (1989)
age estimate (11-12 Myr) appears to hinge primarily on the H-R diagram posi
tion of € Cen, with 5 Cru, a Mus, and Cen defining the rest of the turn-off
isochrone. The latter three stars were confirmed as members by Z99, however
e Cen (the most massive) was rejected. Although our age for e Cen is consis
tent with de Geus's, we omitted it from our LCC age estimate. If one uses the
long-baseline proper motion for e Cen from the new Proper Motions of Funda
mental Stars (PMFS) catalog (Gontcharov et al., 2001), and adopt the LCC space
motion, convergent point, and formulae of Z99 (with vint = 3 km s~'), e Cen has a
100% membership probability. The resulting secular parallax (Tr^ec = 9.6 ± 2.0 mas)
agrees well with the Hipparcos astrometric parallax {TTHIP - 8.7 ± 0.8 mas), further
strengthening the interpretation that e Cen is a bona fide LCC member. Includ
ing e Cen with the other 5 turn-off stars discussed in §2.7.2 does not change our
turn-off age estimate, however (16 ± 1 Myr). If one ignores the stars with masses
less than that of e Cen, then the 12-Myr Bertelli isochrone would appear to be
an acceptable fit for LCC. Because the turn-off is poorly defined, we give equal
weight to the next five Hipparcos members down the mass spectrum (5 Cru, a
Cen, a Mus, Cru, and Cen), which yields an age older than de Geus's.
( Cen is one of several "classical" LCC early B-type member candidates re
jected as members of Sco-Cen using the Hipparcos astrometry. These stars have
been included in Sco-Cen candidate membership lists on and off over the past
half-century: HIP 59196 {5 Cen; B2IVne), HIP 60718A (a^ Cm A; B0.5IV), HIP
62434 {j3 Cm; B0.5IV), and HIP 68702 {/3 Cen; BlIII). These stars are ~ 10-20 Mq-.
96
star, with inferred ages of ~5-l 5 Myr, and distances of ~100-150 pc. Such stars
are extremely rare, and their presense in the LCC region appears to be more than
coincidental. Are they all LCC members whose Hipparcos proper motions are per
turbed due to binarity? All five systems are flagged (field #59) in the Hipparcos
catalog as stars with unusual motions due to either unseen companions or vari
ability. A kinematic investigation of these stars, and their potential membership
in LCC is beyond the scope of this study, but necessary for understanding the
global star-formation history of the Sco-Cen region. Are these stars bona fide
members? If so, why are they so much younger than the other members (both
pre-MS and mid-B stars)? If they are not bona fide members, where did they
come from? Although our age estimates for the pre-MS sample and Hipparcos
early-B members appear to be consistent, the presense of these young, B0-B2 clas
sical members (Hipparcos non-members) hints that the story of star-formation in
Sco-Cen is more complex than our results reveal.
2.8.4 A Star-Formation History of Sco-Cen?
Preibisch & Zinnecker (2000) reviewed the recent star-formation history of Sco-
Cen (<5-10 Myr) in the region of US, UCL, and p Oph. They present evidence
for external supernovae triggering in the formation of the subgroups. They claim
that supernovae shock waves from UCL passed through the US progentior CMC
approximately 5 Myr ago, and caused the cloud to collapse. The US group ap
pears to have had at least one supernova in the past ~1 Myr, possibly a de
ceased massive companion to the runaway 09.5V star ( Oph. This supernova
contributed to destroying the CMC and producing the US superbubble (de Geus,
1992; Hoogerwerf, de Bruijne, & de Zeeuw, 2001). The US subgroup appears to
be currently triggering star-formation in the p Oph cloud core. Here we speculate
97
on the global star-formation history of the Sco-Cen complex.
The formation of LCC and/or UCL may have been similarly triggered. How
ever it is unclear if one triggered the formation of the other or vice versa. Our
pre-MS age estimates are consistent with LCC being slightly older than UCL by a
few Myr, though they could be coeval. What was the origin of these large OB sub
groups? The gas associated with the Sco-Cen complex appears to be part of the
Lindblad Ring, a torus of H I and molecular clouds hundreds of pc in radius. It is
centered roughly near the a Persei cluster and Cas-Tau OB association (Blaauw,
1991; Poppel, 1997). The young stars that have formed from this gas complex (i.e.
the Gould Belt) share a systematic expansion consistent with a localized origin for
the whole complex - possibly an expanding gas shell from a large star-formation
event (e.g. Moreno, Alfaro, & Franco, 1999). The gas associated with Sco-Cen ap
pears to be part of a "spur" of neutral hydrogen and molecular clouds that runs
from near LCC (including Coalsack, Musca, and Chamaeleon clouds), through
Lupus, Ophiuchus, and into the Aquila and Vulpecula Rift regions (see Fig. 3-18
of Poppel, 1997). It is likely that LCC and UCL were among the first clumps
in the Lindblad Ring to collapse and form stars (see §4 of Blaauw, 1991), either
from self-gravity or from triggered from external supernovae events. The LCC
and UCL regions formed a large population of OB stars and their stellar winds
and supernovae may indeed have triggered the collapse of the US group. The
process might continue over the next 10 Myr as the supernovae from the US and
p Oph subgroups send shock waves into the vast reservoir of atomic and molec
ular gas associated with the Aquila Rift (see §3.4 of Poppel, 1997, and references
therein). On the other side of Sco-Cen, there appears to be little gas westward of
LCC until one reaches the Vela complex some 400 pc away. The lack of a suffi-
98
dent gas reservoir probably explains why triggering did not proceed to form OB
subgroups west of LCC.
The Sco-Cen subgroups have formed their own network of superbubbles with
radii of ~100pc (de Geus, 1992). The superbubbles appear to be largely H I, pre
sumably gas from the progenitor Sco-Cen GMC as well as the swept-up interstel
lar medium. In some regions, they are associated with well-known nearby molec
ular cloud complexes: Coalsack, Musca, Chamaeleon, Corona Australis, Lupus,
and numerous small high Galactic latitude clouds (e.g. Bhatt, 2000). The Lupus
clouds are spatially coincident with the western side of the US superbubble, how-
ever no kinematic analysis has been yet undertaken to determine whether the
Lupus clouds share in the bubble expansion. The CrA molecular clouds are em
bedded within the UCL superbubble shell, and the space motion of the T Tauri
star population is moving radially away from UCL (Mamajek & Feigelson, 2001).
Other young stars in the field toward the 4th Galactic quadrant, including the
Tj Cha cluster, TW Hya association, and /5 Pic group, all have ages of ~10 Myr
and are moving radially away from LCC and UCL. Perhaps these stars formed
in small molecular clouds that accumulated within the expanding LCC/UCL su
perbubble shells.
2.9 Conclusions
From our spectroscopic survey of an X-ray- and kinematically-selected sample of
late-type stars in the Sco-Cen OB association, we summarize our main findings
as follows:
99
• We have identified a population of low-mass stars in the Lower Centaums-
Crux (LCC) and Upper Centaurus-Lupus (UCL) OB subgroups with the
following properties: (1) G-K spectral types, (2) subgiant surface gravities,
(3) Lithium-rich, (4) strong X-ray emission (L.v — 10^° -10^' ergs '), and (5)
proper motions consistent with the high-mass members. We classify stars
which show these characteristics as bona fide pre-MS stars or "post-T Tauri"
stars. X-ray and kinematic selection (the RASS-ACT/TRC sample) yielded a
hit rate of 93% for selecting probable pre-MS stars, while kinematic selection
alone (Z99 Hipparcos sample G-K dwarfs and subgiants) yielded 73%.
• We estimate the mean age of the pre-MS population in the LCC subgroup to
be 17 ± 1 Myr (DM97 tracks), 21 ± 2Myr (PSOl), and 23 ± 2Myr (SDFOO).
For UCL, the pre-MS population's mean age is 15 ± 1 Myr (DM97), 19 ± 1
Myr (PSOl), and 22 ± 1 Myr (SDFOO). The UCL pre-MS estimate appears to
be slightly biased towards younger ages (by ~1 Myr) through contamina
tion by Lupus or US members. We also calculate new MS turn-off ages of 16
± 1 Myr for LCC and 17 ± 1 Myr for UCL using the Z99 Hipparcos mem
bership and Bertelli et al. (1994) evolutionary tracks. The UCL pre-MS and
turn-off age estimates are roughly self-consistent, and similar to previously
published estimates. Our age estimates for LCC (pre-MS and turn-off) are
older than previous estimates, and are equal to or slightly older than UCL.
• We find that 68% of the low-mass star-formation in each subgroup took
place within a <4-6 Myr span, and 95% took place within <8-12 Myr (using
DM97 tracks). Hence, we have a strong upper limit of ~10 Myr on the dura
tion of low-mass star-formation in the LCC and UCL progenitor molecular
clouds.
100
Using the Tycho-2 proper motions, we determine that the internal velocity
dispersion of the post-T Tauri population in both LCC and UCL has strong
95% confidence upper limits of 1.6 km s"^ and 2.2 km s"', respectively. The
best estimates of the post-T Tauri population velocity dispersion in LCC
and UCL are cr*„j = O.Glol kms"' and l.Ol?;®, respectively. These values are
consistent with the values (~1 kms~' ) found for the higher mass stars (de
Bruijne, 1999b; Madsen, Dravins, & Lindegren, 2002).
Based on star counts and kinematic arguments, we demonstate that a new
OB subgroup of Sco-Cen proposed to exist in Chamaeleon by Sartori, Lepine,
& Dias (2003) probably does not exist.
We find the frequency of CTTSs among a pre-MS population in an OB as
sociation with masses of 1.3 ± 0.2 (la) MQ and ages of 13 ± 1 (s.e.) ± 6 (la)
Myr (DM97 tracks) to be only 0.9lo;8% (1/110). The younger age results
from using our entire (i.e. magnitude-biased) sample of pre-MS stars. Only
one star in our sample showed both strong Haemission and a K-band ex
cess: the previously known CTTS TYC 9246-971-1 (#34 = PDS 66 = Hen
3-892). This suggests that the disk accretion phase lasts <10-20 Myr in the
evolution of solar-type stars in OB associations.
We demonstrate that a surface gravity indicator for classifying field G and
K stars (Sr 11A4077 to Fe IA4071) can be used to distinguish whether Li-rich
stars are pre-MS or ZAMS in nature. When this indicator is used in tandem
with other youth diagnostics (Li abundance. X-ray emission. Ha emission,
and kinematics), one can confidently classify a star as pre-MS in nature.
101
CHAPTER 3
CONSTRAINING THE LIFETIME OF CIRCUMSTELLAR DISKS IN THE
TERRESTRIAL PLANET ZONE: A MID-IR SURVEY OF THE
30-MYR-OLD TUCANA-HOROLOGIUM ASSOCIATION
This chapter was previously published in Maniajek et al. (2004).
3.1 Introduction
Although the phenonnenon of optically-thick accretion disks appears to be iso
lated to the first '^few Myr of a star's life, numerous examples of older stars with
optically-thin dust disks have been found over the past two decades, primarily
using space-based IR telescopes, e.g. InfraRed Astronomical Satellite (IRAS) and
Infrared Space Observatory (ISO) (Backman & Paresce, 1993; Lagrange, Backman,
& Artymowicz, 2000). Optically-thin disks have been found around stars over
a wide range of ages and masses, but those with the highest fractional luminos
ity (/rf = Lrf,.,jt/L*) are mostly confined to those younger than <fewxlOO Myr in
age (Habing et al., 2001; Spangler et al., 2001). These dusty "debris" disks are in
ferred to be created by the collisions of larger bodies, rather than primordial ISM
dust (Harper, Loewenstein, & Davidson, 1984; Backman & Paresce, 1993). For
micron-size dust grains orbiting between ~0.1-10 AU, the timescale for Poynting-
Robertson drag to pull the grains into the star (~10^~® yr) is short compared to
typical stellar ages (~10^~^" yr), implying that either the observed phenomena is
short-lived, or that grains must be replenished through collisions of larger bod
ies. There is preliminary evidence for a monotonic decrease in dust disk optical
102
depth with age (Spangler et al., 2001), or possibly a more precipitous drop in op
tical depth after age ~400 Myr (Habing et al., 2001). Most of the known debris
disks have been identified by excess far-IR emission above that of the stellar pho
tosphere (e.g. Silverstone, 2000), with characteristic dust temperatures of ~30-
100 K. Despite efforts to find warm (T ~ 200-300 K) dust disks around field stars,
precious few examples with detectable 10-12/i.m excesses are known (Aumann &
Probst, 1991).
Observational constraints on the evolution of circumstellar dust in the terres
trial planet zone are currently scarce. Planned observations with the recently
launched Spitzer Space Telescope (SST) by the Formation and Evolution of Plane
tary Systems (FEPS) Legacy Science program^, among others, will remedy this
situation. FEPS plans to systematically trace the evolution of circumstellar gas
and dust around sun-like stars between the epoch of optically-thick accretion
disks (ages ~ few Myr) to the epoch of mature planetary systems (ages ~ few Gyr;
Meyer et al., 2002,2004).
Although SST promises to provide a leap in our understanding of the circum
stellar environs of stars, we can address a basic question about disk evolution
using currently available ground-based facilities. How much dust remains within
a few Ail of young stars during the epoch of terrestrial planet formation? We address
this question through a mid-IR survey of a sample of young, low-mass stars with
ages of ~30 Myr: the Tuc-Hor Association.
Dynamical simulations suggest that, given the surface mass density of the
minimum-mass solar nebula, runaway growth can take place and form Moon-
sized planetary "embryos" within '-^10® yr (Wetherill & Stewart, 1993). When
the largest embryos reach radii of ~1000 km, gravitational interactions increase
^http://feps.as.arizona.edu
103
the eccentricities and collision velocities of smaller planetesimals, causing more
dust-producing collisions (Kenyon & Bromley, 2004). Over the next ~10^-10^ yr,
the growth of the largest embryos is dominated by giant impacts, which consol
idate the embryos into a small number of terrestrial planets (Agnor, Canup, &
Levison, 1999; Chambers, 2001). During this epoch in our own solar system, the
proto-Earth is hypothesized to have been impacted by a Mars-sized planetesi-
mal, which formed the Earth-Moon system (Hartmann & Davis, 1975; Stevenson,
1987). Chronometry studies using radioactive parent-daughter systems (such as
i82Hf-i82w) suggest that the Earth-Moon impact occurred 25-35 Myr after the for
mation of the solar system (Kleine et al., 2002; Kleine, Mezger, & Mlinker, 2003).
We know that around at least one star (our Sun), terrestrial planets were forming
at age 30 Myr.
In §3.2 we describe the sample, our mid-IR observations, and data reduction.
§3.3 presents the results of our photometry. We present three simple circumstellar
disk models in §3.4, and calculate upper limits to the amount of dust orbiting
within ^10 AU of the stars observed. In §3.5 we discuss our results in light of
previous observational and theoretical efforts in order to better understand disk
evolution around young stars. In the Appendix, we present information related
to the photometric calibration of the MIRAC-BLINC system, as well as details
regardings stars for which mid-IR excesses were detected.
3.2 Observations
3.2.1 The Sample
Our mid-IR survey contains 14 low-mass stars that we argue are probable mem
bers of the ~30-Myr-old Tuc-Hor Association. The observations of the Tuc-Hor
stars are presented in Table 3.1. Observations of some Tuc-Hor candidates that
104
Table 3.1. MIRAC Observations of Tuc-Hor Members
(1) (2) (3) (4) (5)
UT Star Band On-Source Flux
Date Name Time (s) Standards
2001 Aug 8 HIP 105388 N 300 t Cet, a CMa
HIP 105404 N 600 I Cet, a CMa
HIP 107947 N 600 L Cet, a CMa
HIP 108195 N 360 t Cet, Q CMa
HIP 116748AB N 600 i Cet, a CMa
HIP 1481 N 480 t Cet, a CMa
HIP 1910 N 840 h Cet, a CMa
HIP 2729 N 840 L Cet, a CMa
2001 Aug 9 HIP 490 N 420 t Cet, r ] Sgr, a CMa
HIP 6485 N 600 L Cet, r j Sgr, a CMa
I-IIP 6856 N 660 L Cet, Tj Sgr, a CMa
HIP 9892 N 840 t Cet, r i Sgr, a CMa
ERX37N N 1020 t Cet, T) Sgr, a CMa
2001 Aug 10 HIP 9685 N 690 a Cen, t Cet, r ] Sgr
2002 Aug 21 HIP 1910 N 180 7 Cm, t ] Sgr, t Cet
we reject as members, and other young stars (some of which are FEPS SST Legacy
Science targets) are included in Table 3.2. Here we discuss some technical aspects
of the Tuc-Hor sample.
The Tucana and Horologium associations are young stellar moving groups
that were identified nearly contemporaneously by Zuckerman & Webb (2000,
ZWOO) and Torres et al. (2000, TDQOO). Zuckerman, Song, & Webb (2001, ZSWOl)
present an updated membership list and photometry, and suggest that the sim
ilar ages, kinematics, and positions of the Tucana and Horologium associations
allow us to consider them a single group ("Tuc-Hor"). We adopt ~30 Myr as a
reasonable age estimate for Tuc-Hor based on recent values in the literature (see
Table 3.2. MIRAC Observations of Other Stars
(1) (2) (3) (4) (5)
UT Star Band On-Soixrce Flux
Date Name Time (s) Standards
2001 Aug 6 GJ 799 AB 11.6 135 L Get
GJ803 11.6 165 L Get
HIP 108195 11.6 60 (, Get
2001 Aug 7 HIP 99273 11.6 255 a Gar, a PsA, 7 Gru
2001 Aug 8 HD 143006 N 60 t Get, a GMa
HD 143006 11.6 120 7 Gru
[PZ99] J161318.6-221248 N 600 L Get, a GMa
HD 181327 N 240 t Get, a GMa
HR7329 N 60 t Get, a GMa
2001 Aug 9 RXJ1853.1-3609 N 960 t Get, r i Sgr, a GMa
RXJ1917.4-3756 N 660 L Get, r ] Sgr, a GMa
2001 Aug 10 HrP 63797 N 300 a Gen, i Get, r; Sgr
[FZ99] J161411.0-230536 N 360 a Gen, l Get, r j Sgr
ScoPMS 214 N 450 a Gen, i Get, j j Sgr
ScoPMS 5 N 450 a Gen, t Get, 77 Sgr
HIP 95149 N 360 a Gen, t Get, r ] Sgr
HIP 113579 N 510 a Gen, t Get, 77 Sgr
HIP 1134 N 480 a Gen, t Get, r j Sgr
2002 Aug 21 HIP 93815 N 180 7 Gru, 7] Sgr, t Get
HIP 99803 A N 690 7 Gru, r j Sgr, i Get
HIP 99803 B N 165 7 Gru, Tj Sgr, i Get
HIP 105441 N 360 7 Gru, 77 Sgr, t Get
HIP 107649 N 240 7 Gru, T] Sgr, i Get
PPM 366328 N 180 7 Gm, 7] Sgr, I Get
HOP 108809 N 360 7 Gru, r / Sgr, L Get
HIP 108422 N 300 7 Gru, r i Sgr, t Get
GJ879 N 240 7 Gru, J] Sgr, t Get
106
Table 3.3. Age Estimates for Tucana-Horologium Association
(1)
Reference
(2) (3) (4)
Group Age Method
(Myr) ...
(5)
Notes
Zudkerman & Webb (2000) Tuc
Torres et al. (2000) Hor
Torres et al. (2001) Tuc-Hor
Stelzer & Neuhauser (2001) Tuc
Zuckerman, Song, & Webb (2001) Tuc
40 HQ Emission Comparing HQ of 3 members to a Per members
30 Isochrones Siess, Forestini, & Dougados (1997) trades
20 Velocity Disp. "GAYA" = Tuc-Hor
10-30 X-ray Emission Member Lx values similar to TWA, Tau-Aur, & IC 2602
10-40 Isochrones K & M stars; Siess, Dufour, & Forestini (2000) tracks
Table 3.3).
The published membership lists for Tuc-Hor appear to be somewhat subjec
tive and contain some stars that are unlikely to be members. In order to include
stars that are plausibly members of Tuc-Hor as part of our study, we used kine
matics as the primary membership criterion (in addition to the other criteria used
in previous studies). For the combined Tuc-Hor membership lists of TDQOO,
ZWOO, ZSWOl, we calculate membership probabilities using the equations of de
Bruijne (1999a), the heliocentric space motion of the Tucana nucleus from ZWOO,
and the best long-baseline proper motions available at present (preferably Tycho-
2 or UCAC2; H0g et al., 2000a; Zacharias et al., 2003). We reserve rigorous discus
sion of membership and kinematics for a separate future study. For this study, we
retain only those stars whose membership we could not reject based on proper
motion data. We find that ~30% of the stars proposed as members of Tuc-Hor
have proper motions inconsistent with the motion of the assumed Tucana "nu
cleus" (centered on the /5 Tuc mini-cluster; using space motion vector given by
Zuckerman & Webb, 2000). The published membership lists appear to contain
several field stars, a handful of which we observed with MIRAC (Table 3.2) be
107
fore appreciating that they were probable non-members. Care should be used by
investigators employing samples of recently-discovered, diffuse stellar associa
tions (e.g. Tuc-Hor) for the study of age-dependent stellar phenomena.
We include in our Tuc-Hor sample the active dwarf HD 105 (= HIP 490). HD
105 is in the same region of sky (a, d(ICRS) = 00^05™,-41°45') as the other pro
posed Tuc-Hor stars, and has a Hipparcos distance of d = 40 pc. For our calcula
tions we adopt the long baseline Tycho-2 proper motion (H0g et al., 2000a) and
the Tuc-Hor space motion vector from Zuckerman & Webb (2000). In subject
ing HD 105 to the same kinematic tests as the other Tuc-Hor candidates, we are
unable to reject its membership. The calculated cluster parallax (25.3 mas) and
predicted radial velocity (Okms""^) agree very well with the observed trigono
metric parallax (24.9± 0.9mas; ESA, 1997) and measured RV (+1.7±2.5kms~';
Wichmann, Schmitt, & Hubrig, 2003). The equivalent width of the Li I A6707
line (165 niA; Cutispoto et al, 2002) is similar to that for early-G-type members
of ~50 Myr-old IC 2602 and IC 2391 clusters (Randich et al., 2001), and stronger
than that found in ~120-Myr-old Pleiades stars (Soderblom et al., 1993). Finally,
the X-ray luminosity of HD 105 (log(Lx) = 29.4ergs"'; Cutispoto et al., 2002) is
similiar to what is expected for early-G stars with ages of 10-100 Myr (Briceno et
al., 1997). Therefore, we conclude that HD 105 is a likely member of the Tuc-Hor
association.
3.2.2 Data Acquisition
Mid-IR images of the Tuc-Hor members and other young stars were obtained
during the nights of 6-10 August 2001 and 22 August 2002 (UT) with the MIRAC-
BLINC instrument on the Magellan I (Baade) 6.5-m alt-az telescope at Las Cam-
panas Observatory, Chile. The Mid-InfraRed Array Camera (MIRAC-3) contains
a Rockwell HF16 128 x 128 hybrid BIB Si;As array, and was built at the Stew
108
ard Observatory, University of Arizona, and the Harvard-Smithsonian Center
for Astrophysics (Hoffmann et al., 1998). The BracewelL Infrared Nulling Cryo-
stat (BLINC) is a nulling interferometer mated to MlRAC-3 (Hinz et al., 2000). In
our observing mode, however, BLINC is used as a re-imaging system, reducing
the f/11 beam from the Magellan tertiary mirror to a f/20 beam required for the
MIRAC-3 instrument. The pixel scale is 0.123"/pixel, resulting in a FOV of 15.7".
Our intent was to survey for circumstellar dust surrounding our target stars in
the terrestrial planet zone (~0.3-3 AU), with characteristic temperatures of ~300 K,
and corresponding Wien emission peak at ~10 fim. Observations were obtained
with either the wide-band N filter (A^gp = 10.34/mi for an AO star, where Xiso is
the isophotal wavelength, e.g. Golay, 1974) or narrow-band "11.6" filter {Xiso =
11.57/im for an AO star) in standard chop-nod mode (4 position beam switching;
see Appendix 1 of Hoffmann & Hora, 1999). The nod and chop separations were
8", and the chop (frequency of 3-10 Hz) was in a direction perpendicular to the
nod vector. The chop-nod imaging technique produces two positive and two neg
ative images of the star in a square configuration on the detector. The nod separa
tion was chosen so that all four images of the star appear on the detector with suf
ficient room for determination of the background flux in annuli surrounding each
star image. We found that using chop frequencies between 3 and 10 Hz mitigated
the effects of poorly subtracted sky background (background noise increases as
chop frequency decreases) while maintaining observing efficiency (increasing the
chop frequency adds overhead time, with minimal improvement in background
subtraction). Chopping was done with an internal pupil plane beam-switching
mirror within BLINC. The frame time (on-chip integration time) was either 10ms
(jY-band) or 40 ms (11.6-band), and these frames were co-added in 15-30 s long
integrations per nod beam. We found that derotating the MIRAC-BLINC instra-
109
ment (i.e. freezing the cardinal sky directions on the detector) in the Nasmyth
port during observations resulted in poorer background subtraction compared to
turning the derotation off. Hence, for the majority of observations taken during
these nights, the instrument derotation was turned off. The ability to guide the
telescope while derotating the instrument, was not available during our observ
ing runs. Hence the telescope was not guiding during most of our observations.
This had negligible impact on the achieved image quality, but limited our ability
to reliably co-add data for faint sources.
3.2.3 Reduction
The MIRAC images were reduced using the custom program mrclfts (Hora, 1991)
and IRAF^ routines. Flat fields were constructed from images of high (dome) and
low (sky) emissivity surfaces. A median sky frame was produced and subtracted
from the individual (N ~ 10) dome frames. The results were then median com
bined to produce the final flat field. The pixel-to-pixel variation in sensitivity
(~2% r.m.s.) of the MIRAC detector is small enough that flat-fielding had negli
gible effect on our derived photometry.
Aperture photometry was derived using the IRAF phot package. We used
aperture radii of either 0.62" (5 px) or 1.23" (10 px), depending on which flux had
the higher S/N ratio after the photometric errors were fully propagated (dom
inated by sky noise for large aperture or uncertainty in aperture correction for
smaller apertures). The photometry derived with aperture radii of 5 px and 10 px
were consistent within the errors for all of the stars observed. The background
level was determined by measuring the mean sky value per pixel in an annulus
centered on the star with inner radius 1.85" (15 pix) and outer radius 3.08" (25
^IRAF is distributed by the National Optical Astronomy Observatories, which are operated by the Association of Universities for Research in Astronomy, Inc., under cooperative agreement with the National Science Foundation, http://iraf.noao.edu
110
pix) for subtraction. The background annulus radii were chosen so as to sample
a negligible contribution of the star's PSF, but to avoid the PSFs from the other
images of the same star. For the faintest sources, an aperture radius of 5 px was
usually used, in which case an aperture correction was applied to place all pho
tometry on the 10 px system. The aperture corrections were determined nightly
using standard stars, and the typical correction to the 5 px aperture radius pho
tometry was -0.32 ± 0.05 mag. Photometric solutions (zero paints and airmass
corrections) were determined for every night of observations. The typical air-
mass corrections at iV-band were 0.1-0.2 mag airmass"' . The conversion between
fluxes (in mjy) and magnitudes is simply magx{star) = -2.5 log{/A(star)//A(0)},
where /^(star) is the star's flux at wavelength A, and /A(0) is the flux of a zero-
magnitude star (see Appendix D). The sensitivity was such that we could detect
a star with of magnitude 7.5 in iV-band, at S/N ~ 5, with 600 s of on-source inte
gration time.
We observed standard stars taken from the MIRAC manual (Hoffmann &
Hora, 1999), the list of ESO IR standards (van der Bliek, Manfroid, & Bouchet,
1996), and the list of Cohen et al. (1999). Martin Cohen calculated an indepen
dent calibration of the MIRAC photometric system, using the approach identical
to that described in Cohen, Wheaton, & Megeath (2003). Discussion on the in
put data for the absolute photometric calibration, the zero-magnitude attributes
of the MIRAC filter systems, and standard star fluxes, are given in Appendix D.
The absolute accuracy of the standard star fluxes among the four filters ranges
from 1.7-4.5%.
I l l
3.3 Results
/v-band photometry for young stars in Tuc-Hor is presented in Table 3.4, while
photometry for stars in other regions (most belonging to young, nearby asso
ciations) is presented in Table 3.5. Near-infrared {JHK\) photometry from the
2MASS catalog (Cutri et al., 2003) was used to help predict the brightness of the
stellar photospheres at lO/im. For the range of spectral types investigated, mod
els predict that ([11.6] - N) ~ 0.00 within our photometric errors (typically ~0.05-
0.10 mag), hence we plot (A',. — N) and (A',, — [11,6]) on the same color-color plot,
and generically refer to these colors as "(A',, - N)" throughout. A color-color plot
of the Tuc-Hor stars is illustrated in Fig. 3.1.
Table 3.4. AT-band Photometry of Tuc-Hor Members
(1) (2) (3) (4) (5) (6) (7) (8)
Name Name Spec. Ks Pred. F„ Meas. Fi. E(N) Dev.
Type (mag) (mjy) (mjy) (mJy) (o")
HIP 490 HD105 GO 6.12 ±0.02 139 138 ±8 -1±8 -0,2
HEP 1481 HD 1466 F8 6.15 ±0.02 135 151 ±20 +16 ±20 +0.8
HIP 1910 BPM 1699 MO 7.49 ±0.02 51 49±7 -2 ±7 -0.3
HIP 2729 HD3221 K4 6.53 ±0.02 111 102 ±5 -9 ±5 -1,7
HIP 6485 HD8558 G6 6.85 ±0.03 71 82±4 +11 ±4 +2.4
HIP 6856 HD9054 K2 6.83 ±0.02 72 85±4 +13±4 +3.0
HIP 9685 HD 12894 F4 5.45 ±0.02 258 207 ±17 -51 ± 18 -2.9
HIP 9892 HD 13183 G5 6,89 ±0.02 68 67±6 -1±6 -0.2
ERX37N AFHor M3 7.64 ±0.03 42 46±3 +4±3 +1.3
HIP 105388 HD 202917 G5 6.91 ± 0.02 67 65 ±9 -2±9 -0.3
HIP 105404 HD 202947 KO 6.57 ±0.02 92 74±9 -18 ±9 -1.9
HIP 107947 HD 207575 F6 6.03 ±0.02 151 155 ± 12 +4 ±12 +0.3
HIP 108195 HD 207964 F3 4.91 ±0.02 424 394 ±20'>--30 ±21 -1.4
HIP 116748A HD 222259A G6 6.68 ± 0.03 83 75 ±12 -8 ±12 -0.7
HIP 116748B HD 222259B K 7.03 ±0.06 60 53 ±14 -7 ±14 -0.5
Note. — Columns (1) Hipparcos name, (2) other name, (3) spectral type, from
either ZWOO, TDQOO, or ZSWOl, (4) Ks magnitude from 2MASS (Cutri et al., 2003),
(5) measured MIRAC A''-band flux, (6) predicted photospheric flux, (7) flux excess
and uncertainty, (8) residual deviation = E(N)/a{E(N)). ERX 37N is given in
SIMBAD as [TEX22000] ERX 37N. Predicted iV-band photospheric fluxes use or
assume: 2MASS Ks magnitudes. Ay = 0, dwarf color relations from §3.3, and
zero magnitude flux of 37.25 Jy for A^-band.
^•HIP 108195 was also imaged at 11.6^m with a flux of 284 it 40mjy.
113
Table 3.5. Measured Photometry for Other Stars
(1) (2) (3) (4) (5) (6) (7) (8) (9)
Name Name Spec. K. Band Pred.Fj, Meas. Fy E(N) Dev.
Type (mag) (mjy) (mJy) (mJy)
Tuc-Hor Rqects
HD177171 HIP 93815 F7 3.81 iO.W N 1171 1236 ±103 +65 ±149 +0.4
HD191869 HIP 99803 A F7 6.81 ±0.02 N 74 87±9 +13 ±9 +1.5
HD 191869 B'' HIP 99803 B 6.86 ±0.03 N 70 63 ±14 -7 ±14 -0.5
HD 202746 HIP 105441 K2 6.40 ±0,02 N 107 116 ±11 +9 ±11 +0,8
HD 207129 HIP 107649 GO 4.12 ±0.02^^ N 880 837 ±70 -43 ±72 -0.6
PPM 366328 TYC 9129-1361-1 KG 7,61 ±0.02 N 35 63 ±24 +28 ±24 +1.1
HD 208233^ HEP 108422 G8 6,75 ±0.02 N 78 89 ±14 +11 ± 14 +0.8
Upper Sco Members + FEPS
[PZ99] J161411.0-230536 TYC 6793-819-1 KO 7.46 ±0.03 N 48 273 ±11 +225 ±11 +20.3
ScoPMS 214 NTTS 162649-2145 KO 7.76 ±0.02 N 42 37±5 -5±5 -1.0
ScoPMS 5 HD 142361 G3 7.03 ± 0.02'' N 60 58±6 -2±6 -0.4
HD 143006 HBC 608 G6/8 7.05 ±0.03 N 134 648 ±31 +514 ±31 +16,5
HD 143006 HBC608 G6/8 7.05 ±0.03 11.6 104 640 ±34 +536 ±34 +15,7
[PZ99] J161318.6-221248 TYC 6213-306-1 G9 7.43 ± 0.02 N 45 32±6 -13 ±6 -2,2
P Pic Group Members
GJ 799 A HIP 102141 A M4.5 5.70 ±0.10' 11,6 202 259 ±24 +57 ±30 +1,9
GJ 799 B HIP 102141 B M4 5.70 ±0.10' 11.6 202 260 ±25 +58 ±31 +1,9
GJ803 HIP 102409 MO 4.53 ±0.02 11,6 627 608 ±32 -19 ±34 -0,6
HD 181327 HIP 95270 F5/6 5.91 ±0.03 N 169 200 ±18 +31 ± 19 +1.7
HR7329 HIP 95261 AO 5.01 ± 0.03 11.6 301 343 ±31 +42 ±32 +1.3
HR7329 HIP 95261 AO 5 .01 ±0 .03 N 387 456 ±52 +79 ±53 +1.5
114
-| 1 r T 1 r "1 r T r
0.5
0
-0.5 0
• Tuc-Hor O Other
() C)
-i:?
O ( >-
J L J L 1 J L 1 J L J L
0.2 0.4 0.6
(J-KJ
0.8
Figure 3.1: Color-color diagram for Tuc-Hor members (filled circles) and other
stars observed in this study (open circles). Two stars (not members of Tuc-Hor)
with significantly red (KG — N) colors are not shown (HD 143006 and J161411).
The solid line is our adopted photosphere color relation (§3.3). The dashed line is a
smoothed fit to the (Kerr — [12]) dwarf sequence of Kenyon & Hartmann (1995),
where we calculate (K^ — N) ~ {h'cir — [12])KiT95 - 0.06.
115
Table 3.5—Continued
(1) (2) (3) (4) (5) (6) (7) (8) (9)
Name Name Spec. K. Band Pred.Fj, Meas. Fi, E(N) Dev.
Type (mag) (mjy) (mJy) (mJy) (o-)
CrA Off-Cloud Stars + FEPS
RX J1853.1-3609 HD 174656 G6 7.28 ±0.02 N 48 46±4 -2±4 -0.4
RXJ1917.4-3756 SAO 211129 K2 7.47 ±0.03 N 44 45 ±4 +1±4 +0.3
Sco-Cen Reject
HD 113376S HIP 63797 G3 6.70 ±0.02 N 81 83±9 +2±9 +0.2
Other FEPS Targets
HD 181321 HIP 95149 G5 4.93 ±0.02 N 418 425 ±21 +7 ±22 +0.3
HD 191089 HIP 99273 F5 6.08 ±0.03 11.6 113 132 ±15 +19 ±15 +1.3
HD 209253 HIP 108809 F6/7 5.39 ±0.02 N 273 255 ±21 -18 ±22 -0.8
HD 216803 GJ879 K4 3.81 ± 0.02<= N 1233 1027 ±85 -206 ±88 -2.3
HD 217343 HIP 113579 G3 5.94 ±0.03 N 164 160 ±8 - 4 ± 9 -0.4
HD984 mP 1134 F5 6.07 ±0.02 N 145 131 ± 14 -14 ±14 -1.0
Note. — (1) Hipparcos name, (2) other name, (3) spectral type (from SIMBAD imless otherwise
noted) (4) Ks magnitude from 2MASS (Cutri et al., 2003), xmless otherwise noted, (5) measured MIRAC
W^-band flux, (6) predicted photospheric flux, (7) flux excess and uncertainty, (8) residual deviation =
E(N) /a{E iN) ) .
®Star is saturated in 2MASS. We adopt the V magnitude from Hipparcos (ESA, 1997), and the intrinsic
(y — J) and {J — Ks) color for F7 stars from Kenyon & Hartmann (1995) (converted to 2MASS system via
Carpenter 2001), to calculate a rough Ks magnitude. We assume an uncertainty of 0.10 mag.
''Zuckerman et al, (2001) calls the pair HIP 99803 NE and SW. A = SE and B = NW.
'=2MASS photometry is saturated. We take the KCIT magnitude from Aumann & Probst (1991) and
transform it to the 2IvlASS system via equation (12) of Carpenter (2001).
''We can not rule out HIP 108422 as a Tuc-Hor member based on its proper motion, or the agreement be
tween the calcxilated cluster parallax and Hipparcos trigonometric parallax. However, we conservatively
exclude the star as a member at present, since no spectroscopic evidence of youth has been presented in
the literature. If it is co-moving with the Tucana nucleus, we predict a RV of +3 km s~ .
® A 0.8" binary discovered by Ghez, Neugebauer, & Matthews (1993) and seen in MIRAC K-band images.
MKAC N and 2MASS Ks magnitudes are for unresolved pair
f The 2MASS Ks magnitudes from Reid, Kilkenny, & Cruz (2002) appear to be at odds with the combined
magnitude for A & B measured by Cutri et al. (2003), Nelson et al. (1986), and Probst (1983). The system is
essentially an equal brightness binary at optical bands as well as at N, so we split the 2MASS Ks magrutude
evenly and adopt a generous 0.10 mag error. Spectral types are from Hawley, Gizis, & Reid (1996)
sRejected as a Sco-Cen member by Mamajek, Meyer, & Liebert (2002)
116
Fig. 3.1 indicates that { K G —N ) colors are fairly uniform for stars with { J — K G )
< 0.7, i.e. for FGK stars. We decided to analyze the M stars separately from the
FGK stars. We surmise that much of the structure in the published color-color
relations for dwarfs is probably due to statistical fluctuations (e.g. Cohen et al.,
1987; Waters, Cote, & Aumann, 1987; Mathioudakis & Doyle, 1993; Kenyon &
Hartmann, 1995). After examining our data and those from previous studies of
large dwarf samples, we decided to assume a constant (KG — N) color for the
photospheres of FGK stars. In calculating a mean intrinsic (KG — N) color, we
include the FGK stars in Tables 3.4 and 3.5, but exclude the FGK stars HD 143006,
[PZ99] J161411 (both with (/v., — N) ~ 2), and HIP 95270 and HIP 99273 (known
to have far-IR excesses which may contaminate A^-band flux, e.g. Zuckerman &
Song, 2004). The median, unweighted mean, and weighted mean {KG - N) color
for the FGK stars are all similar (0.05, 0.04 ± 0.02, and 0.07 ± 0.02, respectively).
These estimates agree well with the mean FGK dwarf color found by Fajardo-
Acosta, Beichman, & Cutri (2000, {KG — [12]) ~ +0.04 implies {KG — N) ~+0.05)
and are close to the value for AFGK-type dwarfs found by Aumann & Probst
(1991, {KCIT -- [12]) = +0.02 implying {KG - N) = +0.01) Within the uncertain
ties, our measured mean {KG - N) color for FGK dwarfs is consistent with previous
determinations. The mean colors for the young stars in our observing program do not
appear to be biased toward red {KG — N) colors due to the presence of circumstellar ma
terial. We adopt {Kg - N)phot = +0.05 as the photospheric color for FGK-type stars
with (J — Kg) < 0.69.
The observed { K g - N ) colors of the M-type stars are systematically redder
•^Although not explicitly stated, the K photometry from Aumann & Probst (1991) appears to be on the CIT system, where 2MASS KG = KCIT ~ 0.019 (Carpenter, 2001). The MIR AC A'-band photometric system assumes N = 0.00 for Vega, whereas Aumann & Probst (1991) list [12] = +0.01 for Vega. We ignore any color terms in converting [12] to a predicted N magrutude, and derive {Ks-N)~{KCIT-[12])-0.01.
117
than those of the FGK stars, as well as the M-giant standards. We looked for in
dependent confirmation that the turn-up in {K^ — N) color for the coolest dwarfs
is a real effect, and not due to circumstellar material. We measure a color of (A'., —
[11.6]) = 0.33 ± 0.06 for the ^12-Myr-old MO star GJ 803. Song et al. (2002) stud
ied the spectral energy distribution of GJ 803 and concluded that the observed
cold dust excess detected by IRAS at 60/mi does not contribute significant flux at
12/im, and that the IRAS 12//m flux is consistent with a NextGen model photo
sphere. Our observed 11.6//m flux (608 ± 32 mjy) agrees very well with the pre
dicted photospheric flux (633 mJy at ll.bfim), as well as the color-corrected IRAS
12//m flux (537 ± 32 mJy). These observations confirm that the observed {K's -
[11.6]) color for GJ 803 is photospheric, and that the turn-up in f /v', - N) color for
M stars is real. It appears that none of the M-type stars has a statistically signifi
cant mid-IR excess. For the purposes of calculating upper limits on mid-IR excess,
we fit a line to the mean (A', — N) colors for the KM-type stars with (J — A'J > 0.6
(excluding the T Tauri star [PZ99] J161411), and model the M-type dwarf photo
sphere colors as: (A', - N)phot = -0.947 + 1.448x(./ - A',) (0.69 < (J - AT,) < 0.92).
We determine whether a star has detectable excess at A' or 11.6 through calcu
lating the excess as:
E i N ) = N - N ^ n o t - iV - A, + (A, - N),Hot (3.1)
AE{N)] = a^[N] + a''[K,] + a\K, - N\hot] (3.2)
The contribution to the iV-band excess from interstellar extinction will only be
come similar in size to our photometric errors if Ay 1-2 mag, hence we can
safely ignore extinction for the stars observed. We examined the residuals (de
fined as E{N)/a\E{N)]) in order to identify statistically-significant outliers (i.e.
possible A--band excess stars). The ~5-Myr-old Upper Sco members HD 143006
118
and [PZ99] J161411 (Preibisch & Zinnecker, 1999) both stand out with definite N-
band excesses (~ 15-20a), and they are discussed further in Appendix B. There
are two stars with positive 2-3a excesses (HIP 6485 and HIP 6856), however there
are three stars with 2-3a deficits (HIP 9685, GJ 879, and [PZ99] J161318). Hence,
the weak excesses for HIP 6485 and HIP 6856 are probably statistical and not
real. Excluding the two Upper Sco stars with strong A'-band excesses, we find
that 56 ± 12% of the excesses E(N) are within la of zero, and 88 ± 15% are within
2a (uncertainties reflect Poisson errors). It does not appear necessary to intro
duce a non-zero uncertainty in the intrinsic (A',, — N)phot- If one wanted to force
68% of the residuals to be within ±la and 95% to be within ±2a, then either
a[{Ks — N)phot] — 0.07-0.09 mag, or our observational uncertainties are underes
timated by ~40%. We searched for, and could not find, a plausible reason why
our photometric errors would be underestimated by such a large amount. The
observations of our standard stars certainly do not support a significant increase
in our quoted photometric errors. More calibration observations are required to
see if this dispersion can be attributed to actual structure in the intrinsic (A', — N)
colors of normal dwarf stars as a function of spectral type.
Among the Tuc-Hor stars, only one star has as observed A'-band flux >3a
above that expected for stellar photosphere: HIP 6856 (S.OCT excess). However,
there is a Tuc-Hor member with a similarly sized flux deficit (HIP 9685; -2.9a),
so it is difficult to claim that the excess for HIP 6856 is statistically significant.
We find that none of the 14 Tuc-Hor members has an N-band excess more than 3a offset
from the dwarf color relation. We estimate a conservative upper limit to the A'-band
excess due to a hypothetical dust disk as 3x the uncertainty in the flux excess
{a[E{N)\; given in column 8 of Table 3.4).
119
3.4 Disk Models
Our survey was designed to be sensitive enough to detect the photospheres of
young stars, hence we can place meaningful constraints on the census of even
optically-thin circumstellar disks in our target sample. We analyze the upper lim
its to possible mid-IR excess for the Tuc-Hor stars using three different models.
The first model assumes a geometrically-thin, optically-thick disk with a large
inner hole. The second model assumes emission from an optically-thin disk of
single-sized grains. The third model treats the hypothetical disks as a scaled-up
version of the zodiacal dust disk in our solar system.
3.4.1 Optically-Thick Disk With Inner Hole
Infrared and submillimeter observations of T Tauri stars in dark clouds show that
roughly half are orbited by an optically-thick circumstellar dust disk (see review
by Beckwith, 1999). While the stars in our sample are roughly an order of mag
nitude older (~30 Myr) than typical T Tauri stars in dark clouds (<3 Myr), we
can ask the question: If the Tuc-Hor stars have optically-thick, geometrically-thin
disks, what is the minimum inner hole radius allowed by observations?
To answer this question for each star, we adopt the axisymmetric, geometrically-
thin, optically-thick disk model of Adams, Lada, & Shu (1987), and follow the for
malism of Beckwith et al. (1990). While we assume a face-on orientation (9 = 0),
our results are not strongly dependent on this assumption. We also assume that
the disk is optically-thick between ri„ and Vout (300 AU is assumed for all models)
and at all frequencies (1 - ~ 1). This is a safe assumption for T Tauri star disks
in the wavelength regime probed in this study (A < 100 //m; Beckwith et al., 1990).
120
Our A'-band photometry alone allows us to rule out optically-thick disks with
inner hole radii of ~0.1 AU for the Tuc-Hor stars. Stronger constraints on inner
disk radius for a hypothetical optically-thick circumstellar disk can be calculated
by including IRAS photometry. For IRAS point sources, we adopt 25/tm, 60 /im,
and 100 /.i.m fluxes and upper limits from the Faint Source Catalog (Moshir et
al., 1990). Where no IRAS point source is detected, we adopt the IRAS Point
Source Catalog upper limits of 0.5 Jy (25 /im), 0.6 Jy (60 //m), and 1.0 Jy (100 /im)
(IFAC, 1986). For the brightest stars, the IRAS 60 /xm and 100//m data provide
the strongest constraints on the existence of an optically thick disk, whereas for
the fainter K and M-type stars, the MIRAC photometry provides the strongest
constraint. IRAS did not map the region around HD 105, hence the inner hole
radius we derive for this star is based only on the MIRAC iV-band 3a upper
limit. In Fig. 3.2, we illustrate a typical example (HIP 1481) of how the MIRAC
and IRAS photometry constrain the existence of optically-thick disks around the
Tuc-Hor stars. The values we derive for the minimum inner hole radius for a
hypothetical optically-thick disk are given in column 5 of Table 3.6. The median
va lue o f the min imum inner ho le r ad ius i s ~0 .3 AU ( range : 0 .1 -7 .9 AU) . The N -
band and IRAS upper limits place the strongest constraints on inner hole size for
the luminous F stars (vin ^ 5 AU), and the weakest constraints for the faint K/M
stars (rin ^ 0.1 AU). The MIRAC and IRAS photometry easily rule out the existence of
optically-thick disks with inner hole radii of-^0.1 AU of the ~30 Myr-old Tuc-Hor stars.
3.4.2 Optically-Thin Disk
In the absence of circumstellar gas, a putative mid-IR excess around a ~30-Myr-
old star would be most likely to be due to an optically-thin debris disk rather
than an optically-thick T Tauri-type disk. The stellar ages (~10" " yr) are orders
of magnitude greater than the Poynting-Robertson drag timescale (^10^"^ yr) for
121
Table 3.6. Model Parameters for Tuc-Hor Members
(1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11)
Name tt log log tuoie S ri„ - rout So ^diak log Zodys
(mas) TE// L/Lq (AU) (FIM) (AU) (gcm-2) (M®) ( U / U ) { Z )
HIP 490 24.9 3.776 +0.17 0.3 0.50 0.05-13.9 3.4E-06 7.8E-05 -3.25 2.6E3
HIP 1481 24.4 3.780 +0.21 1.8 0.53 0.06-13.8 8.2E-06 1.9E-04 -2.89 6.4E3
HIP 1910 21.6 3.585 -0.81 0.1 0.10 0.03-10.8 1.4E-05 1.9E-04 -2.20 2.9E3
HIP 2729 21.8 3.643 -0.32 0.7 0.27 0.04-12.4 5.0E-06 9.2E-05 -2.80 2.1E3
HIP 6485 20.3 3.744 -0.04 0.3 0.34 0.05-14.4 3.2E-06 7.7E-05 -3.10 2.0E3
H1P6856 26.9 3.707 -0.55 0.1 0.12 0.04-15.3 2.6E-06 7.1E-05 -2.85 1.1E3
HIP 9685 21.2 3.829 +0.71 4.9 1.44 0.09-12.0 6.9E-06 1.2E-04 -3.45 7.5E3
HIP 9892 19.9 3.752 -0.08 0.2 0.31 0.05-14.9 4.5E-06 1.2E-04 -2.91 3.0E3
ERX37N 22.6 3.540 -0.89 0.1 0.14 0.02-8.6 8.3E-06 7.1E-05 -2.48 1.2E3
HIP 105388 21.8 3.746 -0.16 0.2 0.26 0.05-15.3 5.4E-06 1.5E-04 -2.75 3.6E3
HIP 105404 21.7 3.719 -0.21 0.2 0,26 0.04-14.6 6.4E-06 1.6E-04 -2.68 3.8E3
HIP 107947 22.2 3.795 +0.38 2.1 0.75 0.06-13.0 6.3E-06 1.3E-04 -3.16 4.8E3
HIP 108195 21.5 3.817 +0.91 7.9 2.40 0.11-11.7 l.OE-05 1.6E-04 -3.53 8.6E3
HIP 116748A 21.6 3.746 -0.09 1.0 0.31 0.05-14.8 8.0E-06 2.1E-04 -2.65 5.0E3
HIP 116748B 21.6 3.645 -0.46 0.2 0.19 0.04-13.0 1.4E-05 2.9E-04 -2.24 5.8E3
Note. — (1) Star name. (2) Parallax. ERX 37N parallax calculated via cluster parallax method,
the other values are from Hipparcos (ESA, 1997). (3) Stellar effective temperature. (4) Liuninos-
ity in solar units. (5) Lower limits on inner hole radius for a hypothetical optically-thick disk
(§3.4.1). (6) Mean calculated grain size, where a = 5amin/3, where amin is the blow-out grain
size (Eqn. 1; §3.4.2). (7) Inner and outer radii for calculation of opticaUy-thin disk (§3.4.2). (8)
Disk surface mass density for opticaUy-thin disk model (§3.4.2); independent of radius for our
adopted model with (S = So ; p = 0). (9) Upper hmit on disk mass (in grains of size a)
for opticaUy-thin model (§3.4.2). (10) Upper Hmit to fractional luminosity of scaled-up zodiacal
dust model (§3.4.3). (11) Upper hmit to emitting area of scaled-up zodiacal dust model (in units
of "zodys", where IZ = 10^^ cm^; §3.4.3).
122
typical interplanetary dust grains orbiting ~1 AU from a solar-type star. Small
dust grains must be continually replenished by collisions of larger bodies, or else
they would be only detectable for astrophysically short timescales. Using a sim
ple, single grain-size model, we place upper limits on the amount of orbiting dust
within several AU of the ~30-Myr-old Tuc-Hor stars.
Circumstellar dust grains surrounding young main sequence stars should
most likely have radii somewhere between the scale of typical ISM grains (~0.01-
1 //m; Mathis, Rumpl, & Nordsieck, 1977) and solar system zodiacal dust (~10-
100/im; Griin et al., 1985). Theoretically, an ensemble of dust grains produced
from a collisional cascade of fragments is predicted to follow the equilibrium
power-law size distribution (Dohnanyi, 1969): n{a) da = rio da, where p = 3.5.
Indeed this power law distribution is observed for ISM grains (Mathis, Rumpl, &
Nordsieck, 1977) and asteroids (Greenberg & Nolan, 1989). With p = 3.5, most of
the mass is in the largest (rarest) grains, but most of the surface area in the small
est (most common) grains. If the grain size distribution has a minimum cut-off,
the mean grain size is calculated to be a - 5n„,i„ /3 (e.g. Metchev, Hillenbrand,
& Meyer, 2004). A limit on the minimum grain size amin can be estimated from
consideration of radiation pressure blow-out (e.g. Artymowicz, 1988):
3L«Qjrt-l & n G A U c f j
Where L, is the luminosity of the star, Qpr is the radiation pressure efficiency
factor averaged over the stellar spectral energy distribution, G is the Newtonian
gravitational constant, M* is the mass of the star, c is the speed of light, and p
is the grain density (assumed to be 2.5gcm~'^; Griin et al., 1985). The minimum
grain size corresponds to the case where the ratio of the radiation pressure force
to the stellar gravitational force is Frad/Fgrav - 1- For this calculation we assume
123
the geometric optics case where Qpr = 1- For the idealized grain orbiting the Sun
at 1AU, we calculate amin - 0.2 fxm and a = 0.4 //m. The grain size lower limit may
be larger if the momentum imparted by stellar winds dominates radiation pres
sure. The minimum grain size will be somewhat lower if we calculate Qpr using
Mie theory and stellar spectral energy distributions, instead of adopting the geo
metric optics case. Highlighting the uncertainty in this calculation, we note that
the value of a that we calculate for the Sun is ~ 10^ x smaller than the mean in
terplanetary dust particle orbiting in the Earth's vicinity (Grun et al., 1985). This
is largely due to a complex interplay between Poynting-Robertson drag and col
lisions. Increasing the cross-section of dust particles in the solar system zodiacal
dust cloud by ^lO'^ (i.e. comparable to what we are sensitive to in Tuc-Hor, see
§3.4.3), will decrease the collision timescale, and correspondingly decrease the
mean particle size to comparable to the blow-out grain size (Dominik & Decin,
2003).
We model the thermal emission from an optically-thin disk of single-sized
dust grains of radius a orbiting in an annulus between inner radius and outer
radius rout- Spherical grains emit thermally at a temperature 7^ where the inci
dent energy flux from the star is equal to the isotropically emitted output energy
flux of the grain. We approximate the emissivity cx of the single-size dust grains
by using the simple model of Backman & Paresce (1993): emissivity cx = 1 for
A < 'Ittci, and ex = (A/27ra)"^ for longer wavelengths, where we assume (3 = 1.5.
Our adopted value of /5 is similar to that observed for zodiacal dust (see Fig. 2 of
Fixsen & Dwek, 2002) and ISM grains (Backman & Paresce, 1993). The mass opac
ity is calculated as KX — 3exl4:ap. The optical depth of emission through the disk
annulus is TX = where E is the surface density of the disk in gcm~^. We cal
culate the orbital distance from the star (r) of dust grains heated to temperature
124
T using equation #5 from Wolf & Hillenbrand (2003). We verify that this relation
is valid by comparing our calculations with Backman & Paresce (1993) for grains
much larger than the Wien peak of incident light (blackbody case) and for grains
much smaller than the Wien peak of incident light (e.g. ISM grains). For the Tuc-
Hor stars, the Wien peak of incident starlight is comparable to a. We assume a
flat mass surface density profile (E = Eo p — 0), which is appropriate for
a population of dust grains in circular orbits subject to Poynting-Robertson drag
(see discussion in §4.1 of Wolf & Hillenbrand, 2003, and references therein). This
predicted power law is close to what is observed for the zodiacal dust disk in our
own solar system (p ~ 0.34; Kelsall et al., 1998).
Where exactly to define the inner and outer radii of a hypothetical dust disk
requires some basic modeling. Among the Tuc-Hor stars, 90% of the thermal
emission from our hypothetical disk model at N-band comes from within «1.5-
2.2xrw of the star, where rw is the radius at which the dust is at the Wien tem
perature (Tvk) for the isophotal wavelength of the iV-band filter, and Tw =
2898 A^lri (5/(/3 + 5)) (eqn. 6.9 of Whittet, 2003). For simplicity, we adopt a
consistent definition of the outer radius for all stars as 2xrvv Beyond 2x rw, the
hypothetical dust disk contributes negligible flux (^10%) to what is observed in
the iV-band f i l t e r . Approx imate ly 50% of the the rmal emiss ion obse rved in N -
band from a hypothetical dust disk comes from within ^0.4-0.5 x rw of the star.
For rin, we adopt the radius for which the grain temperature is 1400 K - approx
imately the silicate dust sublimation limit. The temperatures of the inner edges
of typical T Tauri star disks appear to be near this value (Muzerolle et al., 2003).
For the Tuc-Hor stars, ~ 0.05 AU and rout ~ 10-15 AU. Hence we are most
sensitive to dust at orbital radii comparable to our inner solar system.
Results for a typical set of fitted model parameters for our optically-thin disk
125
model are illustrated in Fig. 3.2. Our calculations suggest that the survey was
sensitive to dust disk masses of ~ 2 x 10 "^ (~ 10^^ g) in a single-sized dust
grain population (of uniform size a, typically 0.1-2 //m). Our optically-thin model
puts upper limits of So ~ 10~®-10~® g cm"^ on the surface density of micron-sized
dust grains in the ~0.1-10 AU region around the Tuc-Hor stars. For the masses and
surface densities quoted, we assume that all of the mass is in dust grains of size a. In
order to convey how sensitive our assumptions are for our final results, we show
the effects of changing various parameters on our results in Table 3.7. The dust
disk masses that we calculate are similar to those found by other studies (Chen
& Jura, 2001; Metchev, Hillenbrand, & Meyer, 2004) which also use the single
grain-size approximation. The dust mass surface density upper limits that we
calculate are ~10~^-10"® x that of the solids in the minimum mass solar nebula
(Weidenschilling, 1977), however we are not sensitive to bodies much larger than
the wavelength of our observations, or to gas.
3.4.3 Single-Temperature Zody Disk
Another simple model to apply to our data is that of a scaled-up version of the
terrestrial-zone zodiacal dust cloud in our own solar system. Though the detailed
zodiacal dust model for the inner solar system is quite complex (Kelsall et al.,
1998), it can also be approximated by a single temperature blackbody (T = 260 K)
with bolometric luminosity 8 x 10~® Lq (Gaidos, 1999). This luminosity and tem
perature imply an equivalent surface area of 5 x 10~® AU^ ~ 1 x 10^^ cm^. Gaidos
(1999) defines this area as 1 "zody" (1Z). The unit is useful for comparing relative
amounts of exozodiacal dust between the Sun and other stars.
With none of our stars having statistically significant A^-band excesses, we cal
culate upper limits to the number of zodys present using 3x the uncertainty in
the excess measurement, assuming T,i = 260 K, and blackbody emission from
126
Table 3.7. Effects of Changing Adopted Values on Model Output
(1) (2) (3) (4) (5) (6)
Parameter ^^out ASo ^^dust Notes
13 = 2 X (1.00-1.17) x(1.4-2.3) /(1.3-0.26) x (1.5-21) crystalline case
Ii 1—»
/(1.00-1.16) /(1.4-2.3) x (1.2-0.24) /(1.6-22) amorphous case
li o
/(1.01-1.74) /(2.7-12) X (1.5-0.013) /(4.7-10700) blackbody case
S X 10 /(1.01-1.73) /(1.6-4,4) X (9.5-0.32) X (3.8-0.017)
a X 100 /(1.01-1.75) /(1.6-7.1) X (95-2.7) X (38-0.054)
p = 0.34 no change no change x (1.3-0.6) /(1.5-2.8) zodiacal case
p= l no change no change x(1.3-0.17) /(4.7-26)
p = 1.5 no change no change /(1.1-21) / (13-140) min. nxass solar nebula case
Note. — Colximns: (1) model parameter, (2-5) range of factors acted upon model values in
Table 3.6 if parameter in column #1 is adopted, (6) case name.
large grains (i.e. analogous to the situation for the solar system zodiacal dust
disk). An upper limit on the fraction of grain thermal emission to stellar emission
(fd = Ldust/L*) was also calculated for each star using these assumptions. While
the MIRAC photometry is capable of detecting ~4000 Z disks at T = 260 K (equiv
alent log{fd) < —2.9) around the Tuc-Hor members observed, no convincing
mid-IR excesses were detected.
For completeness, we note that the fraction of disk luminosity to stellar lu
minosity {fd = Ld/L*) has been observed to fall off as fd oc (Spangler
et al., 2001). The disk fractional luminosity is predicted to follow fa oc age"^ if
the observed amount of dust is proportional to the collision frequency of large
particles, and P-R drag is the dominant dust removal mechanism. Recently, Do-
minik & Decin (2003) argue that for the very luminous debris disks that have
been detected so far, the collision timescales are much shorter than the P-R drag
127
timescales, all the way down to the blow-out grain size. For this collision-dominated
scenario, Dominik & Decin (2003) predict that the dust luminosity evolves as
f,i (X age "^. While these models ignore effects like, e.g., gravitational perturba
tions, or ejections, of dust-producing planetesimals by planets (which likely had
an enormous effect on the early evolution of the asteroid belt in our solar sys
tem), they provide simple, physically plausible models with which to compare
observations.
If one takes the solar system zodiacal dust disk { l o g { f d ) ~ —7.1, at log{ageyr)
= 9.66), and scale it backward in time according to Spangler et al's relation (/^ cx
^^g-1.76) Qj. thg theoretical P-R drag-dominated evolution (/^ a agt"'^), one would
p r e d i c t a t a g e 3 0 M y r z o d i a c a l d u s t d i s k s w i t h l o g ( f a } ~ - 3 . 3 ( 6 9 0 0 Z ) o r l o g { f d ) ~
—2.7 (23000 Z), respectively. Hence, for the simple model of collisionally replen
ished, P-R-drag-depleted disks, we should have easily detected the solar system's zo
diacal dust disk at age 30 Myr. For the empirical relation (Spangler et al., 2001), we
could have detected the Sun's zody disk around most (13/15) of the ~30-Myr-old
Tuc-Hor stars in our sample. Backward extrapolation of the Sun's zodiacal dust
disk luminosity using Dominik & Decin's relation for collisionally-dominated
disks would yield log{fd) ~ -4.9 (150 Z). Such a disk would not have been
detectable in our survey, consistent with our null result. If analogs of the Sun's
zodiacal dust disk are common around 30-Myr-old stars, must evolve as a
shallower power law (<1.65) than either Spangler et al.'s empirical relation or
the P-R-drag-dominated dust depletion model.
3.5 Discussion
In Fig. 3.3, we plot the incidence of A'-band excess versus stellar age for samples
of low-mass stars. While ~80% of ~l-N'Iyr-old stars have in Taurus have iV-band
128
excesses (Kenyon & Hartmann, 1995), only ~10% of ~10-Myr-old stars in the
TW Hya Association and {5 Pic Moving Group show comparable excess emission
(Jayawardhana et al., 1999; Weinberger et al., 2003a,b). Of these ~10-Myr-old
stars, only a few are known to have optically-thick disks (TW Hya, Hen 3-600),
while the others are optically-thin disks. Our survey imaged a small number
(N = 5) of Upper Sco members (~5-Myr-old) as well, among which two have clear
iV-band excesses. By an age of ~30 Myr, we find that A'-band excesses due to
optically-thick or optically-thick disks, are rare (^7%). Excluding the ~10-Myr-
old star j3 Pic, Aumann & Probst (1991) find only one star ((" Lep) among a sample
of 548 field AFGK stars (?a0.2%) with a convincing 12//m excess. These results
seem to imply that dust appears to be efficiently removedfrom <5-10 All of young stars
on timescales similar to that of the cessation of accretion. While accretion terminates
over a wide range of ages (~1-10 Myr; Haisch, Lada, & Lada, 2001a; Hillenbrand
et al., 2004), the duration of the transition from optically-thick to optically-thin
has been observed to be remarkably short (~10^ yr; Skrutskie et al., 1990; Wolk &
Walter, 1996).
Gravitational perturbations of small planetesimals (—0.1-100km radius) by
growing planetary embryos (~2000 km radius) can theoretically cause collisional
cascades of dust grains that produce observable mid-IR signatures (Kenyon &
Bromley, 2004). Simulations show that when the largest planetary embryos reach
radii of ~3000 km, the population of dust-producing ~0.1-100 km-size planetesi
mals in the planet-forming zone becomes collisional ly depleted, and jV-band ex
cesses become undetectable. The timescale over which the iV-band excess would
be detectable during this phase of terrestrial planet formation is of order ~1 Myr.
With a larger sample size (e.g. FEPS Spitzer Legacy survey), one might be able
to probe whether this is occurring around stars at age ~30 Myr. With 10% of
129
~10 Myr-old stars having detectable A'-band excesses (Jayawardhana et al, 1999;
Weinberger et al., 2003a,b), we may be witnessing the signature from runaway
protoplanet growth in the terrestrial planet zone (Kenyon & Bromley, 2004). This
would agree with the isotopic evidence in our own solar system that Moon- to
Mars-sized protoplanetary embryos accreted within the first ~ 10-20 Myr (Kleine
et al., 2002; Kleine, Mezger, & Munker, 2003), ultimately leading to the formation
of the Earth-Moon system.
3.6 Conclusions
We have undertaken a mid-IR survey of 14 young stars in the nearby ~30-Myr-
old Tuc-Hor Association in order to search for emission from warm circumstellar
disks. No excess emission at 10 /im was detected around any Tuc-Hor members.
If optically-thick disks do exist around these stars, their inner holes must be large
(range: 0.2-5.8 AU). Combining our photometric results with optically-thin dust
disk models, we place the following physical constraints on dust orbiting within
~10 AU of these ~30-Myr-old stars; fractional disk luminosities of L^ust/L^ <
10'^ ® and dust emitting surface areas <4000 x that of the inner solar system zo
diacal dust. The disk masses of micron-size dust grains with orbital radii between
the silicate dust sublimation point and ~10 AU must be less than ~ 10~® M®. The
photometric upper limits also suggest that the upper limit on the surface den
sity of micron-sized grains is ~10' " gcm"'^. These results imply that inner disks
dissipate on timescales comparable to the cessation of accretion.
130
1
0
'-3^
:3
O
- 1
- 2 0 0.5 1 1.5 2 2.5
log(Wavelength) [/xm]
Figure 3.2: Optically-thin and thick dust models fit to the MIRAC and IRAS pho
tometry for a typical Tuc-Hor star (HIP 1481). If the star has an optically-thick
disk, its inner hole radius must be >1.8 AU (constrained by IRAS PSC 60//m up
per limits). The stellar SED is approximated here as a 6026 K blackbody. The
optically-thin disk model is conservatively matched to the 3 x the uncertainty in
the iV-band excess E{N). The kink in the spectral energy distribution for the
optically-thin model occurs at A = lira due to our simple treatment of dust emis-
sivity (§3.4.2).
T I—1—I—I—\—I—I I i—I—I—I—I—I—I—I—I—I—I—I—I—I r
HIP 1481
' Optically-Thin / Disk
I I I I / I I I I I I I I I i l \ I \ I I I I I I L
131
100
80
CO 60 m CD 0 XI
W T3 C S 40
1
20
0 5.5 6 6.5 7 7.5 8
log(age) [yr]
Figure 3.3: TV-band excess versus age for stellar samples of varying ages. Data are
plotted for the following samples: Taurus-Auriga (Kenyon & Hartmann, 1995),
TW Hya Association (Jayawardhana et al., 1999; Weinberger et al., 2003a), ^ Pic
group (Weinberger et al., 2003b), Upper Sco, and Tuc-Hor (both this study). Data
from the IRAS study of AFGK-type field stars by Aumann & Probst (1991) shows
that A^-band excesses among mostly older (;^100 Myr) field stars are extraordi
narily rare (~0.2%). Dashed line is the L-band disk fraction measured by Haisch,
Lada, & Lada (2001a). The solid line is a fit to the A'-band disk fraction for the
four youngest groups. It appears that A'-band excesses are only detectable for
nncirm'nallxr 1 rror -rKanrI cfafic-
Taurus
Upper Sco
TW Hya Assn
(3 Pic MG
Tuc-Hor
132
CHAPTER 4
KINEMATIC DISTANCES TO YOUNG FIELD STARS
4.1 Introduction
4.1.1 Motivation
Often in astronomy, one is interested in studying the nearest objects of a sam
ple. The primary reasons are (1) greatest sensitivity in a fixed integration time,
and (2) the highest possible spatial resolution. Both enable searches for resolved
extended emission (e.g. circumstellar disks) and extremely faint companions at
small physical separations (e.g. brown dwarfs and extrasolar planets).
The measurement of stellar distances is very important for other reasons. A
measurement of distance, along with a determination of the star's bolometric
flux, can be used to infer its luminosity. Combined with theoretical evolution
ary tracks, the luminosity of a star can be used to infer an age and mass. Age
can be especially difficult to estimate through other means. For regions in the
Hertzsprung-Russell diagram where stars evolve relatively quickly (e.g. the pre-
MS Hayashi and radiative tracks), luminosity can be a useful indicator of youth
- especially when combined with corroborating evidence (e.g. stellar activity, Li
abundances). For pre-MS stars in particular, trigonometric parallaxes are rare
(Wichmann et al., 1998; Bertout, Robichon, & Arenou, 1999). In order to un
derstand the recent star-formation history of the solar neighborhood, as well as
identify the nearest, youngest stars for high spatial resolution imaging, we need
reliable methods for assessing distances and ages for field stars.
If one has a model for the velocity field of a stellar group, and can demonstrate
133
that a star plausibly belongs to that kinematic group, then one can estimate the
distance to an object from its proper motion. The calculation of these so-called
secular parallaxes or cluster parallaxes^ has been fundamental to establishing the
cosmological distance scale and understanding stellar evolution. This technique
has been most successfully applied to members of star clusters and associations
(e.g. de Bruijne, 1999b; Madsen, Dravins, & Lindegren, 2002). Theoretically, it
should work for any long-lived over-density in velocity space, whether the stellar
members are born together or not. Olin Eggen used the cluster parallax technique
with stars that he claimed were members of "superclusters" (e.g. Eggen, 1995).
Although the existence of Eggen's superclusters was rather controversial before
his death in 1998 (see Griffin, 1998; Taylor, 2000, for critical reviews of his work),
many of Eggen's groups have been independently found by several studies using
Hipparcos data (e.g. Dehnen, 1998; Skuljan, Heamshaw, & Cottrell, 1999; Chereul
et al., 1999). Among the most reliable identifications are three of the youngest of
Eggen's groups: the Local Association (or "Pleiades Supercluster"), the Hyades
supercluster, and the Ursa Major supercluster (or "Sirius supercluster").
4.1.2 The Local Association
Spectroscopic follow-up of point sources in EUV and X-ray all-sky surveys have
identified a population of Li-rich, late-type stars in the solar neighborhood - pre
sumably young pre-MS or ZAMS field stars. Most of these objects have been
detected in all-sky surveys conducted by the ROSAT satellite at EUV (Pye et al.,
1995) and X-ray (Voges et al., 1999) wavelengths. Many of these young field stars
^In a strict sense, the parallaxes calculated in this work are not moving-cluster parallaxes, secular parallaxes, or statistical parallaxes, as defined in §2.2 of Binney & Merrifield (1998). In a general sense, our method involves two steps: (1) statistically assessing whether a star is a plausible member of a group, and (2) estimating the distance using a predicted tangential velocity and the observed proper motion. Followiiig de Bruijne (1999a), we will simply call parallaxes that are based on a kinematic model and observed proper motion data as secular parallaxes.
134
are more Li-rich than Pleiades members of similar spectral type, so they are likely
to be <125-Myr-old (Stauffer, Schultz, & Kirkpatrick, 1998). The ROSAT All-Sky
Survey contains ~10^ stellar x-ray sources, the majority of which are likely to be
previously unknown, young, active stars (Guillout et al., 1999).
Studies over the past decade reveal that the Li-rich stars found all over the
sky have similar space motion vectors, and a relatively small velocity dispersion
(••^Skms'' ) compared to older field stars (~30 kms~') (Jeffries, 1995; Wichmann,
Schmitt, & Hubrig, 2003; Makarov, 2003). Since these objects were identified in
flux-limited surveys (i.e. X-ray or EUV emission), the kinematic properties of
these objects should be free of kinematic bias. Jeffries & Jewell (1993) and Jeffries
(1995) proposed that the majority of Li-rich field stars found in the ROSAT EUV
all-sky survey are likely to be low-mass members of Eggen's "Local Association"
or "Pleiades Supercluster" (Eggen, 1961,1975,1995,1998). Similarly, Wichmann
& Schmitt (2003) used the X-ray ROSAT all-sky survey to investigate ~750 mostly
G and K-type stars with Tycho proper motions, and found that 10 of the stars
were definitely more Li-rich than the Pleiades (and hence younger). They also
formed a tight kinematic group with a mean space motion similar to Eggen's
Local Association. Wichmann, Schmitt, & Hubrig (2003) suggest that their group
of 10 Li-rich stars could be evidence for the dissolution the Sco-Cen (~15 Myr-
old) or Per OB3 (~50 Myr-old) associations into the field. Regardless of their
origin, it is clear that the nearby, active stars with ages younger than the Pleiades
have remarkably similar space motions within some small velocity dispersion,
and can be considered a kinematic group. Following the work of Eggen, Jeffries,
Wichmann, and other authors, we will refer to this particular concentration^ of
-There is often confusion in the literature regarding the "Local Association" and "Gould Belt." What is commonly referred to as the Gould Belt is a younger (<30-90 Myr) subset of the Local Association which demoastrates peculiar kinematics (a positive K term implying expansion), and appears to be spatially isolated to tilted plane (Lindblad et al. , 1997). The definition of
135
Table 4.1. Parameters for the "Local Association"
Reference Age Uo K, Wo <^u (TV aw Notes
Myr kms~^ kms~^ kms~' kms~^ kms~^ kms~^
Eggen (1998) 0-160? -11.6 -21.0 -11.4 NA NA NA a
Chen et al. (1997) 70 ±40 -10.0 -19,0 -8.1 7.9 8.6 5.8 b
Wichmann, Schmitt, & Hubrig (2003) <125 -11.0 -21.4 -5.5 2.5 3.4 3.0 c
Jeffries (1995) <300 -10 -21 -13 5.7 5.7 5.7 d
^•The age estimate comes from interpretation of remarks in Eggen's papers (1975,1992,1998), including a manuscript ("The
Local Association of Very Young Stars") subnvitted to AJ which went unpublished due to Eggen's death in October 1998.
Eggen claimed that the LA was comprised of young field stars and members of some local clusters, OB and T associations.
Eggen claimed that the oldest duster constituent of the LA was either the Pleiades (125 Myr; Stauffer, Schultz, & Kirkpatrick,
1998) or possibly NGC 2516 (160 Myr; Sung et al., 2002).
''Chen et al. (1997) studied the kinematics of 1,924 BAP-type Hipparcos stars. They found four major dtunps in velodty-age
space, of which the Pleiades supercluster was one.
Wichmann, Schmitt, & Hubrig (2003) surveyed 748 ROSAT-Tycho stars and identified 10 which were more Li-rich than the
Pleiades. Their space motions were remarkably coherent, as listed.
''Jeffries (1995) identified a subsample of ~25 very Li-rich, active, EUV-lxuninous, stars in the ROSAT WFC AU-Sky Survey,
and noted that their kinematic properties were similar to Eggen's LA.
stars in velocity-age space as the "Local Association" (or "LA"), and we list its
vital characteristics in Table 4.1.
We maintain that if a late-type, field star is demonstrably more Li-rich than
the Pleiades open cluster (and hence <125-Myr-old), there exists an excellent,
empirically-measured kinematic model for such objects. We exploit the kinematic
coherence of young stars, and use the observed proper motion vector to derive
secular parallax distances to individual objects. In this chapter, we describe a
technique for calculating secular parallaxes for the instance where a proper mo
tion (/i*, /xa) and radial velocity (p) have been measured for a star at some equa-
the Local Association appears to reflect the fact that the mean space motions of stars in the solar neighborhood younger than 125-160 Myr appear to be reasonably coherent within a small velocity
dispersion.
136
torial position on the celestial sphere { A , S ) , and one has a plausible model of a
stellar velocity field (heliocentric Galactic space motion vector (UQ, Vq, WO), ve
locity dispersion tensor (CU, (TV, (^W), plus Oort parameters {A, B,C,K) to which
the star may belong. The question of whether the star "belongs" to the group, or
should be "rejected" as a field star, can be answered through statistical hypothesis
testing.
In 4.2, we start by reviewing basic vector astrometry, and discuss how cluster
parallaxes are calculated, and how cluster membership is commonly assigned.
We follow this introductory material with §4.3, where we discuss how to general
ize these techniques for velocity fields with anisotropic velocity dispersions and
non-zero Oort constants. We also describe an iterative technique to derive kine
matic distances to field stars that are statistically consistent with membership in
a kinematic group. In §4.4, we discuss our test case (young stars from the FEPS
Spitzer Legacy Science program), and the results (§4.5) for testing the member
ship of these stars to the Local Association. In §4.6, we summarize our findings
and discuss future applications of this technique.
4.2 Astrometry Background Material
4.2.1 Rectangular Equatorial Coordinates
In the rectangular equatorial system, the unit vector x points toward Right Ascen
sion 0 hours (Q = 0°, S = 0°), y points towards Right Ascension 6 hours on the celes
tial equator {a = 90°, 5 = 0°), and z towards the North Celestial Pole (5 = 90°). The
Right Ascensions and declination coordinates are assumed to be on the Interna
tional Celestial Reference System (ICRS), as adopted by the Hipparcos and Tycho
Catalog (see §1.2.2 of ESA, 1997). The ICRS is coincident with the equinox and
equator of the J2000 system (defined by the FK5 catalog) within ~80 mas (Arias
137
et al., 1995), but is tied to several hundred extragalactic radio sources, and is thus
independent of the effects of the Earth's motion. The Hipparcos and Tycho astrom-
etry catalogs are effectively the most accurate realization of the ICRS at optical
wavelengths, and have provided the reference grid for the recent optical/near-IR
sky surveys like 2MASS, Sloan Digital Sky Survey, etc.
The { x , y , z ) vector components are related to the Right Ascension a , declina
tion S, and distance d through the following relations:
X cos (5 cos a
V — d cos 5 sin a
z sin 5
4.2.2 Heliocentric Galactic Coordinates
In our discussion, we use the right-handed heliocentric Galactic coordinate sys-
tern. The unit vector X points toward the Galactic center {i = 0°), vector Y points
towards the direction of Galactic rotation {£ = +90°), and the vector Z points to
ward the North Galactic Pole (6 = +90°), perpendicular to the Galactic plane. The
position of a star in the heliocentric Galactic frame is {X, Y, Z) and its velocity
vector is given by {U, V, W), where U = X, V = Y, and W = Z (dot implies a
time derivative). The position of a star in Galactic coordinates {£, b), with distance
d has a position {X, Y, Z) of:
X cos b cos £
Y = d cos b sin i
Z sin b
The rectangular equatorial coordinates (a:, y , z ) and velocity ( x , i j , z ) can be con
verted to the corresponding Galactic system coordinate {X, Y, Z) and velocity
138
{U, V, W) components through the following matrix relations;
X X X X
F = G y y = G-^ y
Z z z z
U X X v
V = G y y = G-^ V
z z
(4.3)
(4.4)
where G ^ implies the matrix inverse of G and
G =
-0.0548755604 +0.4941094279 -0.8676661490
-0.8734370902 -0.4448296300 -0.1980763734 (4.5)
-0.4838350155 +0.7469822445 +0.4559837762
The elements of the G matrix come from eqn. 1.5.11 of the Hvpparcos and Tycho
Catalog (ESA, 1997) and define the Galactic system on the International Celestial
Reference System (see discussion in §1.5.3 of that reference). The transformation
matrix G is orthogonal, so G~^ is simply equal to the transpose of G (= G^). On
the ICRS, the North Galactic Pole is at ao = 192°.85948 and 6g = +27°.12825, and
the ascending node (where the Galactic equator intercepts the equatorial "Celes
tial" equator) is = 32° .93192. The transformation between equatorial coordi
nates {a, S) and Galactic coordinates can be found in numerous astronomy texts
(e.g. Atanasijevic, 1971), or can be derived from equations 4.1,4.2, and 4.3.
4.2.3 The Normal Triad
One can define a coordinate system known as the normal triad (§1.2.8; ESA, 1997)
with the star at the origin on the celestial sphere, and axes p, q, r (see Fig. 4.1).
139
The vectors p and q are tangent to the celestial sphere, and f is normal to it. Com
ponent p points "east" in positive Right Ascension (a), q points towards "north"
in positive declination (S), and f connects the center of the celestial sphere to the
star, moving positively with increasing radius from the sphere center^. The ve
locity component r corresponds to a star's radial velocity. The tangential velocity
of a star is related to the velocity components in the p and q directions:
V L (4.6)
and the uncertainty in Vtan is:
fp'cX^ 4- (7^(7? rfi(V \ -- ^ P ^ g [^tanj -2 I '2 (4.7)
The velocity vector components { p , q , f ) can be derived from the observed quan
tities
(proper motions //* (= ^aCos{5)), fis, distance d , and radial velocity p ) through
the following relations:
p X
q — y
f z
— sin a + cos a 0
sin S cos a —sinS sin a + cos S (4.8)
+ cos 6 cos a + cos 5 sin a + sin <5
The usefulness of the normal triad system and our definition of Vtan will become
more apparent in §4.2.6. First, we must cover the relationship between space
motion and convergent points, as well as how one tests a star for membership in
a kinematic group.
^This coordinate system is sometimes called the astronomical trihedral in units of d, 5, f (e.g.
Atanasijevic, 1971). We instead adopt the usage of p, q, f, following the convention of the Hippar-cos and Tycho Catalog (ESA, 1997), and some recent influential astrometry papers (e.g. de Bruijne, 1999b; Lindegren, Madsen, & Dravins, 2000)
140
Z (5 = +90")
(a = 90*)
Figure 4.1: The orientation of the normal triad (p, q, f) with respect to the rectan
gular equatorial system (x, y, i) and the celestial sphere. Here, the fiducial star is
at ct = 0° and at some positive declination 5. The p and q axes are tangent to the
surface of the celestial sphere. The rectangular equatorial system is heliocentric,
is defined by International Celestial Reference Frame (ICRF).
4.2.4 Convergent Points and Velocity Vectors
The proper motions (/Xqv fJ-s) for a group of stars with identical space motion vec
tors (x, y, z) will appear to converge towards a convergent point (or vertex). If the
position of the convergent point (acp, Sep) is estimated through another method
(e.g. de Bruijne, 1999a), and the radial velocities (p) for some of the members have
been determined, then one can calculate the magnitude of the group's space mo
tion S:
S = p sec A
cos A = cos Sep cos S cos{acp — a) + sin S^p sin S
(4.9)
(4.10)
141
where A is the angular separation between the star and the convergent point on
the sphere. Conversely, if the space motion and convergent point of the cluster
is known, and the cluster is not expanding or contracting, the radial velocity p
of a cluster member can be predicted as p = S cos A. The velocity vector for a
moving group is related to the convergent point and space motion by the relation:
X cos 5cp cos acp
y = 5 cos 5cp sin acp (4.11)
z sin Sep
If need be, the equatorial velocity vector can be transformed to the Galactic sys
tem via eqn. 4.4. If a published estimate of the space motion vector for a moving
group exists, one can calculate the corresponding convergent point and space
motion:
otcp = tair^ 8cp = tan"' ^ (4.12)
Throughout this discussion, we assume that the velocity vectors of the stellar
members of a kinematic group are normally distributed around the centroid ve
locity vector within some intrinsic velocity dispersion.
4.2.5 Testing for Group Membership
How do we test whether a star "belongs" to a kinematic group? Ultimately, as
signing a star membership to a kinematic group is an exercise in statistical hy
pothesis testing. The best we can do is reject statistical outliers as non-members,
and retain those objects whose properties are consistent with membership as
members. An ideal kinematic group member will have all of its proper motion
pointed toward the convergent point, and none of it perpendicular to the great
142
circle joining the star and the convergent point. We can use this as the basis for
testing the hypothesis of group membership.
The proper motion components {tia-, l^s) can be rotated to see how much of
the stellar proper motion is oriented towards the convergent point and how
much is perpendicular to the great circle between the star and the convergent
point {jjir) '*• Following de Bruijne (1999a), the rotation is calculated as:
+ sin 9 + cos 9
— cos 9 + sin 0
Hv
/ij-
l^a* (4.13)
where the rotation angle 9 is calculated by:
sin (ofcp ~ o;) tan 9 —
cos S tan 6 C P — sin S cos (ttcp ~ C(.)
The proper motion errors are propagated as:
a
2
sin 9 a-\ + cos 6
a:. — cos 9 erf.. Mr fJ'a sin^ 9
(4.14)
(4.15)
(4.16)
The expectation value of fir for a cluster member is zero. Stellar moving groups
have an intrinsic velocity dispersion, and all proper motion determinations will
have measurement uncertainties, so one must determine how close /Xr is to zero
statistically.
In his work assigning membership of Hipparcos stars to nearby clusters and
associations, de Bruijne (1999a) and de Zeeuw et al. (1999) calculate the unitless
quantity t± which measures the deviation of the star's proper motion from that
oriented exactly towards the convergent point.
t± H- F
A/< + *2 int
(4.17)
^We use Kapteyn's original nomenclature of these terms (nvrfJ-r) rather than (/i||,^_L) used by de Bruijne (1999a). The primary motivation is so that we can adopt these letters v and r for unique axis labels which point in the same directions as the proper motion vector ji^)- The notation
(v, f) is arguably more transparent than the cryptic (||,Z)
143
is defined as the ID velocity dispersion of the cluster expressed in proper
motion (masyr~^). In this calculation, the velocity dispersion is assumed to be
isotropic. The quantity is related to the ID velocity dispersion (ajnt) in veloc
ity units (km s"^) by:
where it is the mean parallax of the kinematic group (a good assumption for
spatially compact groups), and A is the astronomical unit (4.74047 km yr s~^).
de Bruijne (1999a) assumes that for a given position on the sky, members of a
cluster will be distributed as a distribution with i/=2 degrees of freedom about
mean values of jUy and /ir- In their analysis, jiy is ignored, and the membership
analysis is totally dependent on the distribution of Hr-
The probability that a true member of the cluster would have a higher than
that observed can be calculated. We generically call this probability {Q) the mem
bership probability. Following the notation of Press et al. (1986), one can calculate
the probability from the complement to the incomplete gamma function;
It is apparent from eqns. 4.17 and 4.19 that stars with low accuracy proper mo
tions (either those with large or small p,r, or both) will have small values of tx
and large membership probability Q. The result is that large numbers of distant
field stars with small proper motions moving in the same direction as a cluster
convergent point, fail to be rejected as members. This was a ever-present prob
lem for selecting members of nearby clusters and associations using Hipparcos
data (e.g. de Zeeuw et al., 1999). While clusters and associations can be isolated
as regions of high stellar density, supercluster members can plausibly be found
f o r u = 2 (4.19)
144
anywhere on the sky. At this point, we do not initially ignore or reject stars with
low S/N proper motions.
In order to assess whether an object is considered a member or non-member,
one must assign in advance a level of significance a. In statistical parlance, a is the
probability of a Type I error, where a hypothesis is true, but rejected by the sta
tistical test (Trumpler & Weaver, 1953). In our case, a is the fraction of bona fide
cluster members that will be accidentally rejected and considered non-members.
Decreasing a too much means that more and more contaminants (field stars) will
not be able to be rejected as non-members, and will be considered members by
the test. Increasing a too much causes more true cluster members to be rejected.
We follow (Trumpler & Weaver, 1953) and adopt a = 0.05 as our fiducial value.
4.2.6 Calculating a Cluster Parallax
The cluster parallax distances for members of moving clusters are some of the
best determined distances in astronomy. The cosmological distance ladder is crit
ically dependent on its first few rungs - the distances to the nearest open clusters.
In this work, we are not interested in determining the convergent point solution
or space velocity for kinematic groups - we rely on previous studies for this in
put. For detailed discussions on how to determine the convergent point for a
stellar group or association, we refer the reader to Atanasijevic (1971) and de
Bruijne (1999a). We are mainly interested in using published convergent point
solutions and high-quality proper motion and radial velocity determinations to
(1) determine whether a given star is a likely member of a stellar group, and (2) if
its membership can not be rejected, estimate its distance. Here we briefly discuss
the cluster parallax technique and present the basic equations. In the following
section we will build upon this technique for use with other velocity fields.
With knowledge of the convergent point and the proper motion of a cluster
145
member, the star's cluster parallax can be calculated:
An All TTc = e • . = (4.20)
S Sin A Vtan
where the constant A = 4.74047 is the astronomical unit (AU) in units of km yr s"' ,
and jj, is the total proper motion {ij? = + //|). The quantity S sin X can be
identified as the tangential velocity Vtan (Eqn. 4.6). In order to extract useful
cluster parallaxes, one needs accurate // values, and work with kinematic groups
which have small intrinsic velocity dispersions compared to S.
4.3 Generalization of the Secular Parallax Technique
With the basic astrometry tools and equations presented in §4.2, one can test a
star for membership in a kinematic group, and calculate a cluster parallax if the
star is a member. In this section, we build on the these techniques, and generalize
the calculation of secular parallaxes for stars which belong to groups with non-
isotropic velocity dispersions, and non-zero Oort parameters. We also formalize
the calculation of membership probability to include radial velocity information.
The motivation for this project is to estimate distances to stars in the FEPS Legacy
Science program, and we are fortunate to be blessed with many radial velocity
determinations for these stars (Hillenbrand, in prep.). Instead of relying solely on
proper motions, we incorporate radial velocity as an another criterion for mem
bership.
This section is outlined as follows. In §4.3.1, we calculate supercluster veloc
ity vectors when the space motion at the position of the Sun is known, as well
as the Oort parameters. In §4.3.2, we generalize the calculation of the velocity
dispersion in the direction tangential to the star-convergent point great circle, in
order to take into account in instance of an anisotropic velocity dispersion tensor.
In §4.3.3 we give a generalized kinematic criterion for supercluster membership.
146
Lastly, §4.3.4 discusses an iterative technique for determining secular parallaxes
to supercluster members. We follow this section with our test case for this tech
nique in §4.4.
4.3.1 Velocity Fields
When calculating a membership probability and cluster parallax for a putative
member of a cluster or association, recent studies assume that the velocity disper
sion is isotropic, and that the mean space motion vector for the group is indepen
dent of position (e.g. de Bruijne, 1999b; de Zeeuw et al., 1999; Madsen, Dravins, &
Lindegren, 2002). For compact open clusters with small velocity dispersion, this
is probably a good assumption. For OB associations, this assumption starts to
breaks down, as the unbound associations are almost certainly expanding (Mad-
sen, Dravins, & Lindegren, 2002), and can cover ~ 10s-100 pc is dimension (de
Zeeuw et al., 1999). On these scales, the effects of differential Galactic rotation
are similar in magnitude to the accuracy of space motion vectors, calculated us
ing Hipparcos astrometry and published radial velocities-^. For characterizing the
space motion of a member of a supercluster, which can be situated anywhere
on the sky, and plausibly anywhere within a few hundred pc of the Sun (due to
the flux limits inherent in one's input sample, here the ROSAT All-Sky Survey
and the Tycho-2 catalog), one can account for first-order corrections to the mean
simple calculation demonstrates this. Assume a hypothetical OB association 100 pc in length, whose space motion matches the Local Standard Rest (on a perfectly circular orbit of the Galaxy). Assume the near side is situated at the Sun, and the far side situated 100pc distant toward (£, b = +90°, +0°). The effects of differential Galactic rotation (assuming A and B Oort constants from Feast & Whitelock, 1997) on the far side will result in a difference of space motion of AU = +2.7km s This is similar in magnitude to the modern accuracy of radial and tangential velocities for members of nearby young groups. The typical errors for published radial velocities for young late-type stars are km s~'. The uncertainty in the tangential velocities for young stars at d ~ 100 pc with Hipparcos astrometry is similarly ~2 km s~ ' (Vtan »: D/i; for accurate H, then (Tvtan/Vtan ~ (ToufDn ~ ao/D ~ 1 mas/10mas ~ 0.1. For Vtan ~ 20kms'^ then crv,„„ ~ 2kmsHence the typical uncertainty in the any velocity component (au, cry, <^w) wi l l typ ica l ly be a t min imum ~ l -2 km s~ ' .
147
superduster velocity vector.
We define the velocity field as a model for the space motion vector of an object
as a function of its position. For a given velocity field, one can calculate the space
motion as a function of Galactic coordinates and distance {£, b,d X, Y, Z) fol
lowing the first-order linear approximation of the velocity field summarized in
§10.3.3 of Binney & Merrifield (1998) and §2.1 of Poppel (1997). As a function of
position (X, Y), the mean U and V velocity components can be corrected using
the Oort parameters:
U
y
Uo +
su Uo
Vo sv Vo
K + C A - B
A + B K ~ C
X
Y
(4.21)
Where the "o" subscript implies that the values are given for the origin, i.e.
the Sun, and corrected for first-order effects like Galactic differential rotation. K
and C are usually statistically consistent with zero for older stellar samples, con
sistent with an axisymmetric Galactic potential. The values of K and C are not
necessarily zero for the youngest stellar samples which may still show irregulari
ties due to the motions of the ISM from which they were born (Torra, Fernandez,
& Figueras, 2000). For our test case (§4.4), we adopt the space motion (C/Q, ^O, WQ)
of the Local Association from Wichmann, Schmitt, & Hubrig (2003), the A and B
Oort parameters from Feast & Whitelock (1997), and assume C = K = 0. In calcu
lating the space velocity as a function of position, we propagate the uncertainties
in the mean Galactic velocity components {Uo, VQ, WO) and the Oort parameters
{A ,B .C .K) .
4.3.2 The Perpendicular Velocity Dispersion Of
The kinematic parameters of superclusters in the literature are nearly always pre
sented in Galactic coordinates. Recall that in calculating membership probabil-
148
ities for stars in open clusters and OB associations (§4.2.5; following de Bruijne,
1999a), an isotropic velocity dispersion tensor was assumed^. The published ve
locity dispersion tensors for superclusters (e.g. Asiain et al., 1999) suggest that
isotropy can be a poor assumption. For superclusters, it would be more appro
priate to estimate the velocity dispersion in the direction perpendicular to the
great circle joining the star and the convergent point (parallel to /x^). We therefore
generalize our membership criterion to reflect the fact that supercluster velocity
dispersions can be anisotropic.
To this end we define (yet) another coordinate system tangential to the celes
tial sphere, which is a simple rotation of the (p, q, f) system. We leave the radial
component f unaltered. This rotation will look familiar as we previously did the
same rotation in transforming the equatorial proper motions (//«., ns) into the
(/i„, /ir) system (i.e. the proper motions towards the convergent point, and per
pendicular to it).
V -1- sin 0 + cos 9 0 P
f = — cos 6 + sin 0 0 q
f 0 0 1 r
The primary purpose of this coordinate transformation is to calculate the quantity
(Tf - the projected intrinsic velocity dispersion of the supercluster in the direction
perpendicular to the star-convergent point line. This quantity is needed for the
revised membership probability calculated in the next section. For clarity, we
® While we have been referring to the published velocity dispersion for kinematic groups (i.e. au,crv,aw) as velocity dispersion tensors, in reality we are only dealing with the diagonal elements of such a tensor. We are currently not tracking covariances in our calculations, and the cross-terms of supercluster velocity dispersions are usually not published. Despite this, we plan on eventually eventually including covariance terms in a future version of the kinematics code.
149
note that an ideal supercluster member would satisfy:
V = = 5- sin A (4.23)
f = 0 (4.24)
r = S cos A (4.25)
i.e. only motion towards the convergent point ( v ) , and in the in the radial direc
tion (r).
4.3.3 Revised Membership Probability
We are now prepared to calculate a revised membership probability for candidate
supercluster members. We incorporate both proper motion and radial velocity
data, as both are readily available for our test case. We assume that for bom fide
supercluster members the observed perpendicular proper motions (as a func
tion of distance), and observed radial velocities p, will be distributed about mean
values (zero and r, respectively) following a distribution with two degrees of
freedom:
The definition for each variable is as follows: Hr (± ) is the observed proper
motion in the r direction (eqn. 4.13), p (± cTp) is the observed radial velocity, and
r (±(Jr) is the predicted radial velocity (and uncertainty) from the model veloc
ity field (from Eqn. 4.8, and simple error propagation of that expression). The
quantity o"| is the intrinsic ID velocity dispersion of the supercluster in the direction
perpendicular to the star-convergent point line at the estimated distance to the star in
proper motion units:
150
where AF is the velocity dispersion in the R direction in kms~'(Eqn. 4.22), TT is
the estimated stellar parallax in milliarcseconds, and A is the astronomical unit
(4.74...) as before. The value is analogous to the term used by de Bruijne
(1999a) for the isotropic ID velocity dispersion of a cluster or association. We
define a* differently from since we are dealing with (1) anisotropic veloc
ity dispersions, and (2) supercluster members do not necessarily clump at some
mean distance, as do members of open clusters or OB associations. The secular
parallax estimate ^ (and uncertainty) is:
« = ~ (4.28) 'tan
^ + (4-29) ^ tan
The membership probability can be derived from the complement to the incom
plete gamma function with u = 2 degrees of freedom (Press et al., 1986), anal
ogous to Eqn. 4.19. Instead of the variable t_i used in Eqn. 4.17 and 4.19, we
intuitively retain use of the value;
Q = —^ r for u = 2 (4.30) r(i//2) V/2
For a star that is not rejected as a member, we estimate the distance D (in pc) from
the calculated secular parallax TT (in mas) from:
B = HE (4.31) TT
4.3.4 Iterative Algorithm
Since the mean velocity vector for a supercluster is position-dependent, and the
distance to the candidate supercluster members is unknown a priori, we devise
an iterative scheme for converging on values of the membership probability and
151
secular parallax for a given star.
[1] Start with an initial guess of the distance (nominally 100 pc - the final
value will be quite insensitive to the initial guess) and calculate the space motion
(U, V, W) from the velocity field parameters (C/Q, 1 O, H Q. A, B, C, K) evaluated at
the star's coordinates {£, b) and initial guess distance d (§4.3.1).
[2] Transform the calculated space motion {U, V, W) to rectangular equatorial
coordinates (x, y, z) (§4.2.2). Calculate the convergent point (cccp, Sep) and space
motion (S) from {x, y, z) (§4.2.4). Calculate the predicted tangential motion Vtan
for a supercluster member at that position (§4.2.3).
[3] Use the convergent point solution {ACP, SEP, S), the predicted tangential mo
tion {Vtan)f and the star's proper motion (//) to calculate a membership probability
and new secular parallax estimate (ff) (§4.3.3).
[4] Using the new secular parallax, repeat steps [1] through [3] until the differ
ence between the old and new secular parallaxes is smaller than some threshold
value.
We find that the secular parallaxes converge to a precision of 10~® within
N ~ 5 ± 1 iterations for the LA convergent point solution. Secular parallax dis
tances are only to be taken seriously for stars whose membership probabilities Q
are above some rejection threshold (a; §4.2.5). Since the membership probabil
ity definition is distance-dependent, and cycling through these calculations on a
modern computer takes a trivial amount of time, we calculate the membership
152
probabilities at each step, but do not reject a star as a member until the final iter
ation.
4.4 Sample Selection
Next we examine the efficacy of our technique for estimating the distances to
young field stars. From a large parent sample (stars from the "Formation and
Evolution of Planetary Systems" (FEPS) Spitzer Space Telescope (SST) Legacy
Science program^ (Meyer et al, 2004)), we select stars that are plausibly <125-
Myr-old by virtue of their Li data, and then kinematically test these stars for
membership in the ^125-160 Myr-old Local Association. The Li selection is an
efficient means of excluding older interlopers from consideration as members.
There are 353 stars in the FEPS program, and the stars were originally se
lected by age, populating six logarithmically-spaced bins between 3 Myr and 3
Gyr. From this parent sample of 353 FEPS stars with extensive ancillary data, we
apply a lithium criteria for isolating those that are plausibly 125 Myr-old. The
majority of the FEPS stars considered have measurements of the equivalent width
of the Li I A6707A line taken at high resolution (R = 20,000) with the Palomar 60"
by Hillenbrand et al. (in prep.). The same study measured radial velocities for
one or more epochs for these stars, with typical accuracy of ~0.6 kms""'. We
omitted most of the young FEPS stars that had reliable membership assignments
(members of Sco-Cen, IC 2602, etc.), but included some as "tracers" as a check of
our technique.
We placed statistical constraints on the ages of the FEPS stars by comparing
their Tg// and EW(Li) data to that for the 125-Myr-old Pleiades cluster. First,
we combine all of the high-resolution measurements of EW(Li) for Pleiades FGK
^http: /,/leps.as.arizona.edu
153
stars from the literature (Butler et al., 1987; Pilachovvski, Booth, & Hobbs, 1987;
Boesgaard, Budge, & Ramsay, 1988; Soderblom et al, 1993; Garcia Lopez, Re-
bolo, & Martin, 1994; Jones et al., 1996; Jeffries, 1999). If multiple measurements
were available for any star, we preferred the values taken at higher resolution
or higher signal-to-noise. Overall, 153 Pleiades members had high resolution
EW(Li) measurements. Three stars (HII948, HIT 1794, HII2208) were rejected as
Li-poor outliers, and were noted as such in previous studies. We fit a polynomial
to the data in the region devoid of EW(Li) upper limits, between 4300 K ^ ^e// <
6720 K, where we have 119 data points. The distribution of log(EW(Li)) values for
Pleiades members is well represented by a quadratic fit with different variances
on either side of 5300 K.
l og{EW{Li ) ) = -5.962278 + (3.133483£^ - 3) x - (2.986343£; - 7) (4.32)
Where x = log(EW(Li)), the EW(Li) values are in units of mA, and their published
error estimates are typically ~0.02 dex. The observed rms scatter for Pleiades
members between 4300 K < T,.yy < 5300K is 0.213 dex, and only 0.066 for those
with 5300 K < Te// < 6720 K.
We classify stars as younger than the Pleiades when their EW(Li) value is
higher than 95% of Pleiades for a given effective temperature. For a normal dis
tribution, 5% of Pleiades members will have EW(Li) values more than +1.645cr
above equation 4.32. In other words, with data for bona fide Pleiades cluster
members, we would consider 95% as being consistent with the age of the Pleiades,
and 5% would fall above +1.645(7, and be incorrectly considered younger than
the Pleiades. Indeed, testing the 119 Pleiades members in the temperature range
where our fit is relevant, we find that 4/119 (3.4%) are rejected as Li-rich out
liers. This is within the Poisson noise, and consistent with our prediction. Our
154
, , 1 , 1 1 1 1 1 <125 Myr
1 1 1 1 1 1 1 -old region
' 1 '
-
. * % • \
•• \ • •
M M M '
K -
t -
K * *7 •
•X M
• « » s « "
w * -/%• * •
• • ^ -
M _
- M I I I . i l l ! ) , , , 1 , , ,
i» -, 1 ,
7000 6000 5000 4000
Te„/K
Figure 4.2: The distribution of Tg// vs. EW(Li) values for Pleiades members. The
region for selecting stars that we consider <125-Myr-old is outlined in a dashed
line.
test will, however, reject many stars whose ages are likely to be <125 Myr, but
whose EW(Li) values lie in the locus of Pleiades data points. For example, a
~50-Myr-old Li-poor star may have a EW(Li) value within the locus of points for
the 125-Myr-old Pleiades. Such a star would not be selected as younger than the
Pleaides.
Of the 353 stars in the FEPS Legacy Science survey, 287 (81%) currently have
measurements of the Li I A6707A line. Of these 287 stars, 72 have lithium values
within the "<125 Myr" selection region in Fig. 4.2, i.e. are more Li-rich than 95%
of Pleiades for a given 11//.
Of these 72 Li-rich stars, we chose roughly half (32) that had accurate ra
155
dial velocity measurements (required for our kinematic analysis). Most of these
remaining stars that were previously known to be members of clusters or as
sociations were omitted, but a few were retained for independent verification.
Hence, our sample consists of ^10% of the original FEPS sample, but includes
the youngest stars in the program. The input astrometry and photometric data
for these stars are listed in Table 4.2. As mentioned in §4.3.1, we adopt the Local
Association as defined by Wichmann, Schmitt, & Hubrig (2003, parameters given
in Table 4.1) as the supercluster to which we test membership for the young FEPS
stars.
Table 4.2. Sample of FEPS Targets Younger than the Pleiades (<125 Myr-old)
Name a(ICRS) 5(ICRS) fi<5 '^trig Spec. Te// EW(Li) RV Av Prob, Kin.
hms o ' " mas yi~ mas yr~ mas Type K mA kms~^ mag % Mem,
RE J0137+18A 01:37:39.41 +18:35:33.16 +65.8 ±1.9 -46,0 ±2,5 K3Ve 4837 416 ±7 +1.8 ±2.2 0.00 72,3 Y
1RXSJ025751.8+115759 02:57:51.68 +11:58:05.83 +31.4 ±1.2 -28,4 ±1.2 G7V 5509 232 ±12 +12.0 ±2,2 0.72 ±0.08 41.9 Y
IRXS J031644.0+192259 03:16:43.89 +19:23:04.11 +11.2 ±1.3 -8.5 ±1.3 32±44 G2V 5856 190 ±10 +8.2 ±1.2 0.04 ±0.07 35,6 Y
IE 0324.1-2012 03:26:22.05 -20:01:48.81 +25.0 ±1.6 +7.4 ±1,6 -74 ±39 G4V 5717 203 ±13 +26.0 ±0.6 0.07 ±0.05 0,0 N
RXJ0331.1+0713 03:31:08.38 +07:13:24.78 +13.7 ±1.5 -13.1 ± 1,3 K4Ve 4632 371 ± 11 +15.4 ±0.3 0.00 34,1 Y
HD 22179 03:35:29.91 +31:13:37.45 +42.6 ±0.6 -46,0 ±0,7 15 ±11 G5IV 5986 179 ±9 +11.6 ±0.6 0.12 ±0.10: 4.5 N
1RXSJ035028.0+163121 03:50:28.40 +16:31:15.19 +26.2 ±1.3 -23,4 ±2.1 G5IV 5681 206 ±11 +8.0 ±0.6 0.37 ±0.06 66.8 Y
RXJ0357.3+1258 03:57:21.39 +12:58:16.83 +22.7 ±1.8 -21.9 ±1.5 -20 ±98 GO 5970 245 ±16 +6.8 ±0.6 0.58 ±0.06 31.4 Y
HD 285281 04:00:31.07 +19:35:20.70 +2.7 ±1.1 -12.9 ±1.2 -37 ±40 K1 5123 409 ±9 +16,1 ±0.6 0.69 ±0.05 0.1 N
HD 285372 04:03:24.95 +17:24:26.12 +4.8 ±2.0 -14.4 ±2.0 K3V 4759 653 ±13 +10,2 ±0.7 0.78 ±0.12 15.6 Y
HD 284135 04:05:40.58 +22:48:12.14 +6.0 ±0.6 -14.9 ±0.6 16 ±19 G3V 5793 208 ±8 +12.8 ±1.2 0.09 ±0,04 4.8 N
HD 279788 04:26:37,40 +38:45:02.37 +5,8 ±1.1 -28.6 ±1.0 114 ±67 G5V 5681 267 ±14 +11.4 ±0.4 0,68 ±0,07 0.1 N
1RXSJ043243.2-152003 04:32:43.51 -15:20:11.39 +2,3 ± 1.1 +14.2 ±1.1 80 ±35 G4V 5779 278 ±12 +17.6 ±0.6 0,37 ±0.05 0.0 N
RXJ0442.5+0906 04:42:32.09 +09:06:00.86 +28.9 ±2.4 -22.3 ±2.0 132 ±52 G5V 5661 247 ±12 +19.7 ±0.6 0.45 ± 0.09 10.8 Y
HD 286179 04:57:00.65 +15:17:53.09 -1.8 ±1.5 -17.3 ±1.4 -13 ±31 G3V 5705 205 ± 15 +13,8 ±0,4 0.41 ±0.06 0.0 N
HD 31950 05:00:24.31 +15:05:25.28 +0.3 ±1.1 -15.2 ±1.1 -41 ±31 6104 150 ±8 +13,9 ±0.6 0,12 ±0,10: 0.5 N
HD 286264 05:00:49.28 +15:27:00.68 +20.0 ±1.4 -59.0 ±1.4 K2IV 4995 423 ±7 +18,1 ±0.6 1,09 ±0,08 19.5 Y
1RXSJ051111.1+281353 05:11:10.53 +28:13:50.38 +6.0 ±0.8 -24.0 ±0.7 36±64 KOV 5267 464 ±9 +11,6 ±1,2 0.92 ±0,07 37.3 Y
HD 245567 05:37:18.44 +13:34:52.52 +7.5 ±0.9 -33.2 ±0.9 26 ±22 GOV 6050 276 ±9 +14,4 ±0,5 0,63 ±0,04 44.8 Y
SAO 150676 05:40:20.74 -19:40:10.85 +19.2 ±1.2 -12.9 ± 1.2 4±10 G2V 5827 216 ±8 +24,8 ±0.5 0,00 95.7 Y
AO Men 06:18:28.24 -72:02:41.56 -7.9 ±1,0 +74.3 ±1.0 26±1 K3V 4359 357: +16.2±0.7 0,00 20.4 Y
157
4.5 Results
In subjecting the astrometric and radial velocity data for the young FEPS stars
to our iterative procedure in §4.3.4, we found that 24/32 (75%) of the objects are
statistically consistent with Local Association membership. Since the stars were pre
selected to be more Li-rich than the Pleiades, we already know that their true
ages (as inferred from Li abundances) are probably consistent with being Local
Association members.
With the availability of color-magnitude information, reddening estimates,
and effective temperatures, we would like to plot these stars on the HR diagram
to see if they are consistent with being very young, i.e. pre-MS. In order to do this,
we calculate how much above the ZAMS a particular star is using both empirical
isochrones and theoretical evolutionary tracks. It has been recently found that
no set of widely-used, modern isochrones can fully fit the observed color-magnitude data
for open clusters across the full main sequence (Grocholski & Sarajedini, 2003). The
systematic differences between the different theoretical ZAMS lines (in {V - /v,)
vs. My) are of the order A{V - Kg) ~ 0.05 - 0.1 mag, which can translate into
luminosity offsets of AlogL/L© ~ 0.05-0.1 dex for solar-type stars. To mitigate
against the systematic differences between theoretical tracks and observed color-
magnitude data, we do the following:
[1] Using the available V and A', photometry, and reddening estimate Ay from
the FEPS database (Carpenter, in prep.), we deredden the observed V magnitude
and {V - Kg) color to produce Vo and (V - A',)o-
[2] Estimate the absolute magnitude My for a ZAMS star of similar {V — Kg)o -
Table 4.2—Continued
Name a (ICRS) 5(ICRS) lla* Spec. Te// EW(Li) RV Av Prob. Kin.
hms o ' " mas yr" masyr~i mas Type K mA kms~^ mag % Mem.?
RX J0850.1-7554 08:50:05.41 -75:54:38.11 -18.5 ±1.5 +33.1 ± 1.3 -14 ±27 G5 5711 250 ±15 +15.5 ±2.0 0.31 ± 0.02 80.1 Y
HD 86356 09:51:50.70 -79:01:37.73 -25.9 ±1.1 +39.0 ±1.5 68 ±19 G6/K0 5611 260 ±15 +12.2 ±2.0 0.52 ±0.03 10.8 Y
RX Jllll.7-7620 11:11:46.32 -76:20:09.21 -17.2 ±5.2 +6.9 ±5.2 K1 4621 500: +19.0 ±2.0 1.30 ±0.05 17.2 Y
HD 104467 12:01:39.15 -78:59:16.85 -41.0 ±1.0 -4.4 ±1.1 2±7 G5in/IV 5690 255: +21.7 ±4.0 0.14 ±0.02 6.9 Y
HD 141943 15:53:27.29 -42:16:00.81 -42.2 ± LI -65.3 ±1.1 8±13 G0/2V 5805 230: +1.9 ±5.6 0.00 62.0 Y
HD 142361 15:54:59.86 -23:47:18.26 -29.3 ±1.1 -38.8 ±1.1 6±15 G3V 5833 260 ±7 -4.6 ±0.8 0.45 ±0.02 42.8 Y
ScoPMS 214 16:29:48.70 -21:52:11.91 -5.6 ± 3.6 -22.1 ± 1.8 KOIV 5318 426 ±11 -7.6 ±0.9 1.69 ±0.03 12.1 Y
HD 174656 18:53:05.99 -36:10:22.91 +3.8 ±1.3 -24.7 ±1.1 -63 ±29 G6IV 5629 348 ±6 -5.1 ±0.6 0.84 ±0.04 25.2 Y
liX J1917.4-3756 19:17:23.83 -37:56:50.52 +8.3 ± 1.3 -27.3 ±1.0 37 ±31 K2 5001 443 ±9 -2.4 ±0.6 0.49 ± 0.05 6.1 Y
HD 191089 20:09:05.22 -26:13:26.63 +39.3 ± LI -68.2 ±1.2 19±1 F5V 6441 96 ±5 -7.8 ±2.2 0.00 17.8 Y
HD 202917 21:20:49.95 -53:02:03.05 +30.2 ±1.5 -97.2 ±1.6 22±1 G5V 5553 233: -5.3 ±1.0 0.00 5.3 Y
Note. — Colximns: (1) Name, (2) Right Ascension (ICRS), (3) declination (ICRS), (4-5) proper motion in R.A. and dec, from UCAC2 catalog if available (Zacharias et
al., 2003), or Tycho-2 (H0g et al., 2000a), or Hipparcos (ESA, 1997), (6) trigonometric parallax from Hipparcos or Tycho-1 (ESA, 1997). Those with errors in parallax of >5
mas are from Tycho-1, the rest from Hipparcos. Columns (7-11) are from the "Formation and Evolution of Planetary Systems" (FEPS) Spitzer Legacy Science database.
(7) spectral t5fpe (mostly from Michigan Spectral Atlas Vols. 1-5; e.g. Houk & Cowley, 1975), (8) effective temperature (Carpenter et al., in prep.), (9) equivalent width of
Li I A6707.7A(Hillenbrand et al., in prep.), (10) radial velocity (Hillenbrand et al., in prep.), (11) V magnitude extinction Ay (Carpenter et al., in prep.), (12) Membersfiip
probability Q defined in §4.3.3, (13) Is the star selected as member of the Local Association based on kinematic criteria (Q > 5%)?
159
[3] Calculate the difference in absolute magnitude (AMy) between that pre
dicted with the secular parallax distance (My(sec)) vs. that predicted with the
empirical ZAMS (My(ZAMS)).
[4] Estimate the difference in bolometric luminosity between the secular par
allax solution and the ZAMS solution as Alog{L) — -0.4 A My. To calculate a
luminosity "calibrated" for use with the evolutionary tracks, add Alog{L) to the
luminosity of an object with similar Te/f on the 625 Myr-old solar metallicity
isochrone.
For the empirical ZAMS, we adopt the high-quality, single-star ZAMS for the
Hyades from Pinsonneault et al. (2004) ((V^ — Ks)o vs. My). We then corrected
this ZAMS to solar metallicity using the {V - A',) vs. [Fe/H] correction from
Sarajedim et al. (2004), and assuming the Hyades metallicity ([Fe/H] = +0.13)
from Paulson, Sneden, & Cochran (2003). The color shift to solar metallicity was
A{J — Kg) = -0.024 mag. For the solar metallicity 625-Myr-old isochrone, we inter
polate values from the evolutionary tracks of D'Antona & Mazzitelli (1997), and
normalize these tracks to the Hyades locus for solar metallicity.
Since our stars were selected to be < 125-Myr-old by virtue of their lithium, we
predict that the HRD for these stars should mostly consist of ZAMS stars, with
some pre-MS stars. The pre-MS epoch for a 1 MQ stars is ~40 Myr, so if the sam
ple stars were equally distributed in age between 0-125 Myr, we would naively
expect ~2/3rds to be ZAMS, and 1 /3rd to be pre-MS, and no stars below the
ZAMS. In order to estimate an isochronal age for each star, we take the estimated
T^ff and "calibrated" logL/L.,, values, and generate 10' Monte Carlo HRD po
sitions for each star with the associated uncertainties in and criog(LfLQ)- We
160
interpolate ages and masses for these Monte Carlo points on the D'Antona &
Mazzitelli (1997) evolutionary tracks. Due to the results of our lithium test in
§4.4, we rejected those Monte Carlo-generated ages and masses when the isochronal ages
wer >125 Myr. We also reject isochronal ages of <0.1 Myr, as these HRD positions
would lie above the stellar birthline and be inconsistent with the appearance of
an unembedded star with little reddening. The fraction of the Monte Carlo data
points that are rejected due to our Li criterion is listed in column (8) of Table
4.3. For the remaining Monte Carlo values of age and mass, we calculate the me
dian value and assign 68.3% (la) confidence intervals. These values are given in
columns (6) and (7) of Table 4.3. The HRD for the FEPS stars that we select as
Local Association members is shown in Fig. 4.3.
As seen in Fig. 4.3, none of the FEPS stars selected as Local Association mem
bers is more than 1.2a below the ZAMS. Some of the stars appear to be very
young (<10 Myr). We see that roughly half of the stars (11) are within 2a of
the ZAMS. Only three of the stars in Table 4.2 have accurate Hipparcos distances.
The predicted secular parallax distances versus the Hipparcos distances are as fol
lows: AO Men (37 ± 1 pc vs. 50 ± 9 pc; +1.4a), HD 191089 (54 ± 3 pc vs. 59 ± 9 pc;
+0.6a), and HD 202917 (46 ± 2 pc vs. 53 ± 7pc; +1.0a).
Can the isochronal ages listed in Table 4.3 be independently verified by other
means? AO Men is claimed to be a member of the ~12 Myr-old j5 Pic Moving
Group (Zuckerman, Song, Bessell, & Webb, 2001). Using the iterative secular par
allax calculated for AO Men, we derive an isochronal age for AO Men of llig"
Myr, in agreement with Zuckerman et al.'s assessment. HD 202917 is claimed
to be a member of the ~30 Myr-old Tucana-Horologium association (Zuckerman,
Song, & Webb, 2001, we agree with this assessment in §3 of this thesis). The itera
tive secular parallax calculated for HD 202917 leads to an isochronal age of 28
161
Table 4.3. Kinematically-Selected Local Association Members
Name Distance AMv logCTi//) logL/L© Isochronal Age Mass Fraction
pc mag K log(yr) Mo %
REJ0137+18A 64±8 -1.14 ±0.31 3.685 ±0.005 -0,20 ±0,12 g 97+0-22 -0.21 1 ^•^^-0.09 100
1RXSJ025751,8 118±16 -0.11 ±0.45 3.741 ±0.004 -0,18 ±0,18 7 ^R+0.17 100^®'^^ 67
1RXSJ031644.0 365 ± 60 -1.30 ±0.50 3.768 ±0.004 0,53 ±0,20 l,42t°;?| 99
RXJ0331.1+0713 250 ±40 -3.12 ±0.38 3.666 ±0.005 0,45 ±0,15 c cc+0.23 0.00_0 19 0.801;°:°® 100
1RXSJ035028.0 138 ±21 -0.24 ±0.41 3.754 ±0.004 -0,01 ±0.17 7 Q0 + O.1O 1-061^:^5 72
RXJ0357.3+1258 149 ±23 0.54 ±0.45 3.776 ±0.004 -0,12 ±0,18 7 +0-06 ' •^•'•-0.10 1.15 ±0.02 13
HD 285372
RXJ0442.5+0906
343 ±69
119 ±21
-2,78 ±0.52
0.72 ±0.58
3.678 ±0.005
3.753 ±0.004
0,39 ±0,21
-0,41 ±0,23
c qc+0.30 O.yO-o 27
7 43+0-15 '•^^-0.14
i.UO_Q 11
03
100
12
HD 286264
1RXSJ051111.1
71 ±11
199 ±29
-0.51 ±0.44
-2.36 ±0.40
3.699 ±0.004
3.722 ±0,004
-0,35 ±0,18
0,55 ±0,16
'••^^-0.27 6 44+^-20 "•^^-0.21
0 92"'"°'^® "•^'^-o.io 1 00+0.22
21
90
100
HD 245567
SAO 150676
AO Men
RXJ0850.1-7554
119 ±21
78 ±30
50±9
112 ±18
-0.58 ±0,42
-0.07 ±0.89
-1,14 ±0,38
0,23 ±0,35
3,782 ±0,004
3,765 ±0.004
3.639 ±0.005
3.757 ±0.004
0,37±0,17
0,02 ±0,36
-0.49 ±0,15
-0.18 ±0.14
7 oq+0.14 ' •^'5-0 12
' "^'^-0.24
' •'-'^-0.34
' -^^-0.09
-^•^"-0.08 1 1 7+0.25
' -0.08
0.841^7
91
53
100
28
HD 86356
RXJllll.7-7620
99 ±15
291 ±91
-0,30 ±0,34
-4,01 ±0,71
3.749 ±0.004
3.665 ± 0.005
-0.03 ±0.14
0.80 ±0,28
7 07+0.10 '"^'-0.13 c qc:+0.26 O.C5O-0 21 0,84l°;l«
82
84
HD 104467
HD 141943
118 ±15
67±9
-1.89 ±0,29
-1,04 ±0,29
3,755 ±0,004
3,764 ± 0,004
0,65 ±0,12
0,39 ±0,12
"•'^-0.17
7.12±0.11
1.68l°;f,
1 qO + 0.11 ^•^•^-0.12
100
100
HD 142351 101 ± 14 -1.08 ±0.32 3,766 ±0,004 0,42 ±0,13 7.10 + 0.12 1.34 ±0.12 100
ScoPMS 214 215 ±41 -2.45 ±0.42 3,726 ± 0,004 0,62 ±0,17 6.41 ±0.21 l,94i°;2® 100
HD 174656
RXJ1917.4-3756
193 ±29
173 ±25
-2.40 ±0.37
-2.72 ±0.34
3,750 ±0.004
3,699 ±0,004
0.81 ± 0.15
0.52 ±0.14
GAStlit 2,08j:°;23«
j.,jo_o 10
100
100
HD 191089
HD 202917
59±9
53 ±7
-0.13 ±0.34
-0.05 ±0.28
3,809 ±0,003
3,745 ± 0,004
0.47 ±0.14
-0.17 ±0,12
''^^-0.12
7.44 ±0.16
1 -^•^^-0,07 0 QQ+0 ®'^
68
66
Note. — Columns (1) Name, (2) distance from secular parallax estimate, (3) difference in absolute magnitude between
unreddened star at secular parallax distance and ZAMS, (4) effective temperature (Carpenter, in prep.), (5) luminosity of
star calculated according to description in §4.5, (6) median value of log(age/Myr) and 68.3% confidence limits from HRD
position on D'Antona & Mazzitelli (1997) tracks (assuming Gaussian errors in logTe// and logL/Lg), (7) median value of
mass and 68.3% confidence limits from HRD position on D'Antona & Mazzitelli (1997) tracks (assuming Gaussian errors in
logTe// and log L/L©), and (8) percentage of Monte Carlo points (N=10^) which gave reasonable isochronal ages between
0.1-125 Myr, and were included in the estimation of ages and masses for columns (6) and (7).
162
2
1
0
•1
3. 3.75 3.65
Figure 4.3: HRD for FEPS stars selected as Local Association members. The lumi
nosities were calculated using secular parallax distances following the discussion
in §4.5. Evolutionary tracks are from D'Antona & Mazzitelli (1997) and a 1998 up
date for <1 MQ stars.
Myr, in agreement with the mean age for its assigned group. Two stars in Table
4.3 have been previously classified as members of the ~5-Myr-old (Preibisch et
al., 2002) Upper Sco OB subgroup (HD 142361 and ScoPMS 214). The iterative
secular parallax distances lead to isochronal ages of 1313 Myr and 31? Myr for
HD 142361 and ScoPMS 214, respectively. We conclude that for the instances
where FEPS stars in Table 4.3 have been previously classified as members of
other groups, our technique leads to isochronal ages consistent with published
estimates of the group ages.
A few stars in our subsample appear to be extremely young (<1 Myr), but
163
not embedded in known star-forming clouds. Can we truly believe that some
of the FEPS stars (RX J0331.1+0713, HD 285372, RX Jll 11.7-7620) have ages of <1
Myr? Two effects may be conspiring to produce what might seem to be an unusu
ally high number of < 1-Myr-old stars. First, one of the sources for FEPS targets
was spectroscopic surveys of ROSAT All-Sky Survey X-ray sources in the vicin
ity of star-forming molecular clouds. The first two (RX J0331 and HD 285372) are
south of the Taurus-Auriga clouds, where isolated Classical T Tauri stars have
been found (Wichmann et al., 1996). The third source (RX Jll 11.7-7620) is actu
ally on the outskirts of the Chamaeleon I cloud. While these stars have not been
kinematically linked to these star-forming clouds, it is plausible that they were
recently formed in molecular clouds related to Taurus or Chamaeleon (within the
past <few Myr). So we can not dismiss these very young age estimates easily.
Note that we have also ignored stellar multiplicity. RX J0331 has been resolved as
binary at K-band (Palomar 200" AO survey; Metchev, Hillenbrand, & White, in
prep.) with the companion being roughly half the brightness of the primary. HD
285372 is a single-lined spectroscopic binary candidate (Wichmann et al., 2000).
However, RX J1111.7 appears to be single according to multi-epoch radial veloc
ity monitoring (Melo, 2003) and a speckle search for companions (Kohler, 2001).
Hence, the first two stars may be somewhat older stars (~10 Myr) whose unre
solved binarity (according to our Tycho-2 and 2MASS photometry) makes them
appear younger on evolutionary tracks. Unresolved binarity among bona fide
stars may be responsible for scattering some stars into the pre-MS part of the
HRD, but it is doubtful that is responsible for all of the high luminosity stars in
Fig. 4.3.
164
4.6 Conclusions & Future Work
For the few instances where Local Association stars in the FEPS database have
previously published distances and age estimates, we find satisfactory agreement
between our values and those previously published. For many of the stars in
Table 4.3, verification (or negation) of these secular parallax estimates will have
to await the next astrometric space mission that will conduct an all-sky survey:
GAIA (Ferryman, 2003).
The tools developed in §4.2 and 4.3 can be used as a powerful engine to iden
tify the nearest, youngest stars in the solar neighborhood. If one cross-references
a deep proper motion catalog (like the UCAC2 catalog; Zacharias et al., 2003) and
an all-sky X-ray survey (i.e. ROSAT; Voges et al., 1999), one can identify active
stars whose proper motions are consistent with membership in the Local Asso
ciation. The Local Association velocity field is appropriate for the young B-type
population (Chen et al., 1997) and the Li-rich field stars (Wichmann, Schmitt, &
Hubrig, 2003), and so provides an excellent kinematic model with which to com
pare the motions of candidate young stars.
Although a previous survey of ROSAT-Tycho stars with statistically signifi
cant trigonometric parallaxes yielded only ~10 <125-Myr-old stars out of a sam
ple of ~750 targets (Wichmann, Schmitt, & Hubrig, 2003), this search was biased
against identifying pre-MS K and M stars due to the flux-limit of the Tycho-
2 proper motion catalog {V < 12""'-'). The recently available UCAC2 catalog
(Zacharias et al., 2003) will allow us to probe three magnitudes deeper than Tycho-
2, and should allow kinematic selection of the nearest K and M-type pre-MS stars
within ~50 pc of the Sun. Spectroscopic follow-up of hundreds of candidates will
be necessary to confirm whether the X-ray active stars co-moving with the Local
Association are indeed pre-MS. We conclude that it is now practical to search for
165
the nearest, isolated pre-MS stars in the solar neighborhood, without having to
wait 1-2 decades for the next dedicated space-based astrometric mission.
166
CHAPTER 5
CONCLUSIONS & FUTURE WORK
The stated goals of this thesis were to investigate the fossil record of star-formation
in the nearest OB association and to study the evolution of circumstellar environ
ments within 10 AU of young solar-type stars. In order to do this, we needed
to develop new techniques necessary for identifying elusive post-T Tauri stellar
populations. We have succeded in achieving these goals, and in the process cre
ated a tool which can be used for the next generation of surveys to identify the
nearest, youngest stars to the Sun.
We identified a large sample of nearby post-T Tauri stars in the nearest OB
association to the Sun (Sco-Cen). Our study of this young, solar-type popula
tion in §2 found that the duration of star-formation in the LCC and UCL OB
subgroups was short (<5-10 Myr), consistent with the idea that star-formation
in giant molecular clouds takes place over timescales shorter than the ambipo-
lar diffusion time scale (>10 Myr). Hence, one does not expect large numbers
of ~10-30 Myr-old post-T Tauri stars on the periphery of modern star-forming
clouds.
The answer to the question "Where are the post-T Tauri stars?" is this: Post-
T Tauri stars are found in ghostly kinematic groups devoid of obvious molec
ular clouds. They can be found as low-mass members of OB associations (like
Sco-Cen), and in kinematic groups that escaped detection until recently (e.g. rj
Cha cluster, TW Hya association, Tuc-Hor association). The idea that post-T
Tauri stars may be found in "halos" surrounding long-lived molecular clouds
167
that are currently forming stars, appears to lack observational support. Molec
ular clouds appear to form stars during brief bursts (<few Myr), followed by
disruption of the cloud by the new stellar population, and termination of further
star-formation. In the 20th century. Classical T Tauri stars (CTTS) were discov
ered in large quantities by objective prism surveys of star-forming clouds, and
weak-lined T Tauri stars were discovered in X-ray surveys of those same clouds.
In the 21st century, we are now capable of finding the long-sought post-T Tauri
stars in large numbers using all-sky X-ray and proper motion datasets.
As a by-product of our spectroscopic survey of pre-MS stars in Sco-Cen in
§2, we found that the incidence of accretion disks among -^1 MQ stars in a stel
lar population with mean age ~13 Myr is ~1 %, indicating that it is very rare
for young solar-type stars to exhibit Classical T Tauri star activity beyond this
age. Building on this work, in §3 we surveyed a group of post-T Tauri stars
whose age corresponded with the epoch of terrestrial planet formation in our
solar system (Tucana-Horologium association; age ~30Myr). We found no indi
cation of warm dust emission from circumstellar material within ~10 AU of these
young stars. Through modeling of the flux upper limits with plausible circum
stellar disk and dust grain models, we conclude that these post-T Tauri stars have
less than ^3000 times the dust grain surface area of our solar system's zodiacal
dust disk. If the stars have optically-thick dust disks, then our photometry rules
out inner hole radii of <0.2-6 AU. Combining our survey results from other re
cent surveys suggests that detectable --^10 /imemission from circumstellar disks
(either optically-thick T Tauri accretion disks or optically-thin debris disks like
that orbiting /5 Pic) disappears by age —20 Myr. Theoretical models of terrestrial
planet-formation suggest that dust is generated collisionally in copious amounts
by perturbations from Ceres-sized planetesimals. The effects are observable as
168
mid-lR excess emission compared to the stellar photosphere. This dust gener
ation mechanism dies off within a short time (~1 Myr) after the planetesimals
reach this threshold size. Those models, combined with observations from our
survey and other published surveys, suggest that planetesimals reach ~1000 km
in radii before ~20 Myr age. At the age of the Tuc-Hor association (~30 Myr),
the oligarchical accretion of the largest planetesimals has probably swept up the
hoards of smaller, dust-generating bodies.
In §4, we developed a revised method of assigning membership and calculat
ing kinematic distances (from an iterative secular parallax technique) to putative
members of superclusters. Our results show that the isochronal ages that we de
rive using these distances are consistent with previously published ages. Hence,
we have developed a technique for identifying pre-MS field stars that currently
lack trigonometric parallax information. One could, in principle, apply these
techniques to identify the nearest, youngest stars in the solar neighborhood. The
basic ingredients for such a search are available today: the ROSAT All-Sky Sur
vey can identify active stars by virtue of their X-ray activity, and the Tycho-2 or
UCAC2 proper motion catalogs provide photometry and proper motion infor
mation. Using this input data, and employing a reliable kinematic model for the
<100-Myr-old stars in the solar neighborhood (i.e. the Local Association), one
can statistically test whether the stars have motions consistent with membership
in the group. Spectroscopic follow-up of the kinematically-selected stars would
be necessary to identify those that are Li-rich, and to measure radial velocities,
which can be used to further refine the probability of membership. Finding the
nearest, youngest, stars is crucial for advancing our knowledge regarding the for
mation and evolution of planetary systems, and for placing our own solar system
in context.
169
APPENDIX A
THE PRE-MAIN SEQUENCE Te// SCALE
Pre-main sequence stars lie between dwarfs (V) and subgiants (IV) on color-
absolute magnitude or temperature-luminosity H-R diagrams. A given visual
spectral type will correspond to cooler temperatures as surface gravity decreases
(e.g. Gray, 1991; de Jager & Nieuwenhuijzen, 1987). Dwarf temperature scales are
often adopted for pre-MS populations, however it is prudent to account for the
effects of surface gravity.
We quantify the effects of log g on the SpT vs. T^ff relation using two datasets.
First, we fit a polynomial surface to Te//(SpT, log g) using the data from Gray
(1991, Table 2). As with most compilations of T^//(SpT) in the FGK-star regime,
we find that a trinomial is the best low-order fit, and that a linear dependence
on log g adequately accounts for the effects of surface gravity on temperature.
Our second method finds a similar surface fit to TE//(SpT, log g) using published
Teff and log g values (Cayrel de Strobel et al., 2001) for the GK standards of
Keenan & McNeil (1989) and F standards of Garcia (1989) (those within 0.3 dex of
solar [Fe/H], and luminosity class IV and V only). We adopt the isochrones from
D'Antona & Mazzitelli (1997) for a fiducial log g value as a function of Teff for a
coeval 15-Myr-old population.
Both assessments yield essentially the same result: dwarf temperature scales
for G-K stars should be lowered by 35 K for a 15 Myr-old population. The tem
perature decrement increases for younger ages: 70 K-40 K for G0-K2 10-Myr-old
stars, 235 K-105 K for G0-K2 5-Myr-old stars, and 260 K-180 K for G3-K2 l-Myr-
170
old stars. Both techniques yielded a linear dependence of log g on Te//(SpT,
logg), and the slopes were similar: dTeff/dlogg ~ 220 K/log(cm s~^) for the
Keenan standards with Cayrel de Strobel stellar atmosphere data, and dT,.///91 og g
~ 190 K/log(cm s~^) for the interpolation of Gray's (1991) Table 2. As the evo
lutionary model isochrones are parallel to the main sequence when the stars are
on the radiative tracks (i.e. ~ 10-30 Myr for rsj IMe stars), the Alog g between a
15 Myr-isochrone and the main sequence is fairly constant over the G-K spectral
types. Hence one naively expects a linear offset in Teff{SpT, log g) between the
15-Myr isochrone and the MS.
With the l a scatter between published T,,//(SpT) relations being ~60 K amongst
G stars, the systematic shift is nearly negligible. Upon comparing several tem
perature scales from the literature, we adopt the dwarf T(.f/ scale from Schmidt-
Kaler (1982), and apply a -35 K offset to correct for the effects of lower surface
gravity for a putative 15-Myr-old population. We conclude that adopting dwarf
7].FF vs. SpT scales for pre-MS stars younger than ~10 Myr will systematically
overestimate their T^f/ values, and in turn, their masses inferred from evolution
ary tracks. This could have deleterious systematic effects on derived initial mass
functions for young associations.
171
APPENDIX B
STANDARDS WITH QUESTIONABLE LUMINOSITY CLASS
Several of the standard stars we observed had H-R diagram positions, published
logp values, and Sr II A4077/Fe I A4071 ratios (SrFe) which differed from what
is expected for their luminosity classes given in Keenan & McNeil (1989). The
differences are only at the half of a luminosity-class level. We adopt the Keenan
temperature types for ail of his standard stars, however we revised the luminos
ity classes of these stars to bring their H-R diagram positions, SrFe index, and
published log g estimates into harmony (Table B.l). The SrFe indices for the vast
majority of the standards formed loci according to luminosity class (Fig. 2.1), so
we are comfortable using the index as an additional descrinninent. The dwarf re
gression line in the gravity indicator vs. temperature indicator plot (Fig. 2.1) was
constructed using only Keenan standards for which his luminosity classification
agreed with published log g values and the H-R diagram position.
172
Table B.l. Revised Luminosity Classes of Standard Stars
Star Published log q Sr/Fe HRD adopted
Lum. class estimates class pos. Lum. class
HR 5072 V 3.8-3.9 (IV-V) IV-V IV-V IV-V
I IK 4995 IV-V 3.0-3.7 (IIl/IV) IV IV IV
HR 5409 IV 3.3-3.9 (III/IV) III-IVIII-IV III-IV
HR 6608 Illb ... IV IV IV
Note. — Published luminosity classes from Keenan
& McNeil (1989). The range of published log^ esti
mates come from the compilation of Cayrel de Strobel
et al. (2001). "Sr/Fe class" is from measuring the Fe I
A4071/Sr II A4071 ratio in our spectra and intercompar-
ison to the spectral standards in Table 2.3 (see Fig. 2.1).
The luminosity class from the HRD position uses the V and
(B - F)data from it Hipparcos and the standard relations
from Appendix B of Gray (1991).
173
APPENDIX C
POLYNOMIAL FITS
• MI6 v s . Spec t ra l Type : This flux ratio is Index 6 of Malyuto & Schmidt-Kaler
(1997). We measure the index in magnitudes (MI6 = -2.51og(f(/\A5125-5245) /f(AA5245-
5290))), and find the following relation for Keenan and Garcia F0-K6III-IV stan
dards (Table 2.3) within 0.3 dex of solar metallicity:
+0.06 < M/6 < 0.26 :
S p T = 33.26 ± 0.07 + (22.75 ± 0.48) x M/6
-0.04 < M/6 < 0.06 :
SpT = 30.82 ± 0.16 + (105.38 ± 8.28) x M/6 - (711.74 ± 173.81) x M/6^ (C.l)
where SpT is the spectral type on Keenan's (1984) scale, i.e. F5 = 28, F8 = 29, GO =
30, G2 = 31, G5 = 32, G8 = 33, KG = 34, K1 = 35, and K2 = 36. Intermediate types
can be assigned e.g. G9 = 33.5, G9.5 = 33.75, K0+ = 34.25, K0.5 = 34.5, etc. The first
equation applies to K0-K6 stars, and the second equation applies to FO-KO stars.
The la dispersion in these fits is 0.6 subtypes.
• X4374/X4383 vs. Spectral Type: This band ratio consists of two 3A bands
centered on 4374.5A and 4383.6A. We measure the index in magnitudes as YFe =
-2.5xlog(f(A4374.5)/f(A4383.6)). We find the following relation between the band
ratio and SpT for F0-K6 Ill-V stars;
= 25.97 ±0.30+(20.47 ±0.90) x YFe (C.2)
174
The residual standard deviation to the fit (using 20 Keenan F0-K5III-V standards
within 0.3 dex of solar metallicity) is 0.6 subtypes.
• Surface Gravity Index Fe I X4071/Sr II X4077 vs. Spectral Type: We measure
a surface gravity index using the flux ratio of two SA bands centered on Fe I
A4071.4 and Sr II A4076.9. We measure the flux ratio in magnitudes: SrPe = -
2.5 X log(f(A4071) /f(A4077)), and plot against our MI6 spectral type index. The
Keenan standard dwarfs confirmed as being main sequence stars define a narrow
locus:
SrFe = -0.078 ± 0.005 + (2.123 ± 0.261) x MI6 - (8.393 ± 2.945) x MI6'^
+ (15.487 ±8.672) x MI6^ (C.3)
The la sample standard deviation of this fit is 0.0094 mag in SrFe. The boxmdary
between dwarfs and subgiants in Fig. 2.1 is -2a of the dwarf locus. This relation
is valid for F9-K6 stars.
• EW(Ha) vs. Spectral Type: In Fig. 2.2, we fit the equivalent widths of the Ha
feature (at low resolution, the photospheric absorption plus the chromospheric
emission) as a function of spectral type for F0-K6 dwarf and subgiant standard
stars (Table 2.3) with the following polynomial:
E W { H a ) = 2.983 ± 0.066 - (0.456 ± 0.027) x ( S p T - 30)
+ (2.574 ±0.378) x lO"'-^ x {SpT - 30)^
EW(Ha) is measured in A. SpT is spectral type on Keenan's scale (as before). The
sample standard deviation of the polynomial fit to 11 standards was 0.20A.
• Converting Tycho [B — V) to Cousins-Johnson B-V: The Hipparcos catalog
gives linear relations between (BT — Vr), {B — V), V, and Vy for stars of a wide
range in spectral types. Bessell (2000) compared the Hipparcos/Tycho photometry
175
and that of the E-region photometric standards, and refined the relations between
the two systems. Table 2 of Bessell (2000) gives a standard relation between {BT —
VT) Cousins-Johnson (B - V) and (V - VT) for B-G dwarfs and K-M giants. We fit
the following relations to Bessell's tables;
= V r + 9.7 X K) -' - ] .334 x U r ^ B r - V T) + 5.486 x 10"^(/ir - V r f (C.5)
— 1.998 x 10 '~(B'I• —
{ B - V ) = { B T - V T) + 7.813 x U )-'\B T - V T) - 1.489 x {B T - V T)^
+ 3.384 x 1(}~^{BT — VT)^ (C.6)
(S - y) = {BT - VT) ~ 0.006 - 1.069 x lO-^^r - VT) + 1.459 x 10-^(5t - VT)^
(C.7)
The Y{VT,{BT - VT)) polynomial equation C.5 applies to stars from -0.25 <
{BT - VT) < 2.0 (B-M types). Equation C.6 is for stars with 0.5 < (BT - VT) <
2.0, and equation C.7 is for stars with -0.25 < {BT — VT) < 0.5. We do not quote
uncertainties in the polynomial coefficients since the Bessell relations are already
smoothed. These equations fit Bessell's standard relations to 1-2 millimagnitudes.
176
APPENDIX D
MIRAC PHOTOMETRIC CALIBRATION
With a significant body of MIRAC-BLINC observations acquired during the 2001-
2003 observing runs at Magellan I, it was decided to calculate the photomet
ric attributes of commonly used MIRAC bands on the Cohen-Walker-Witteborn
(CWW) system of absolute infrared calibration (e.g. Cohen, Wheaton, & Megeath,
2003, and references therein)'. The photometric standard system for previously
published MIRAC studies is given in Appendix 2 of the MIRAC3 User's Manual
(Hoffmann & Hora, 1999).
Relative spectral responses (RSRs) for each combination of filter and window
were constructed. The throughput chain consists of the following groups of com
ponents: atmosphere, telescope optics, BLINC optics, MIRAC optics, and MIRAC
detector. The complete throughput equation consists of the following compo
nents multiplied together: atmosphere, 3 aluminum mirrors (Magellan), dewar
window (KRS-5 or KBr), KBr lens (in BLINC), 5 gold mirrors (3 in BLINC, 2 in
MIRAC), filter, and the Si:As array. Most of the MIRAC observations were taken
in just four of the seventeen filters currently available in the three MIRAC filter
wheels: L, N, 11.6, and and these were the filters we absolutely calibrated. For
each MIRAC filter, we list a manufacturer's name, mean filter wavelength (Ao),
'This appendix was previously published in Mamajek et al. (2004) so as to provide an absolute photometric calibration of the publicly-available MIRAC camera for the community. The author completed the work regarding digitizing archived printouts of filter transmission profiles and gathering optical throughput measurements, however Martin Cohen is credited with calculating the zero-magnitude attributes and absolute photometric calibration provided in Tables D.l and D.2. Many thanks go to Phil Hinz and Bill Hoffmann for providing critical information that made this calibration possible.
177
bandwidth (AA/A; defined as the FWHM of the normalized transmission curve
divided by the mean wavelength), and the temperature at which the filter profile
was measured (or extrapolated). The transmission profiles for the filters are plot
ted in Fig. D.l. While we list mean filter wavelengths (Aq) in this discussion, the
isophotal wavelengths (Xiso) are given in Table D.l.
The L filter (OCLI "Astro L"; Ao = 3.84/;/m; AA/A = 16.2%; 77 K) is the same one
used in all previous and current MIRAC L-band observations, and the transmis
sion curve is plotted in Fig. A2.2 of Hora (1991). The N filter (OCLI code W10773-
8; Ao = 10.75 //m; AA/A = 47.2%; Ambient) was purchased in 1994 in preparation
for comet Shoemaker-Levy-9 observations, and has been in use ever since. Pre-
1994 MIRAC observations employed a slightly bluer wideband N filter whose
characteristics we only present here for completeness (OCLI code W10575-9; Ao =
10.58 //m; AA/A = 45.8%; Ambient). The narrow 11.6 fim filter (OCLI "Astronomy
R"; Ao = 11.62//m; AA/A = 9.5%; extrapolated to 5 K) has been used since MIRAC
was commissioned (Hora, 1991). Its transmission curve includes the effects of
a BaF2 blocker, and it is the only filter of the four that we were able to linearly
extrapolate its transmission characteristics to the detector's operating tempera
ture (5 K; data at Ambient and 77K were available). The Q., or "Q-short" filter
has also been used for the lifetime of MIRAC, and its characteristics are only cur
rently known at ambient temperature: Aq = 17.50 //m; AA/A = 10.6%).
We followed Cohen et al. (1999) in using PLEXUS (Clark, 1996) to assess mean,
site-specific atmospheric transmission. Transmission curves for KRS-5 and KBr
were taken from the Infrared Handbook (Wolfe & Zissis, 1985). We used a KRS-5
window during the August 2001 and May 2002 runs, and a KBr window for the
August 2002 and March 2003 runs. The reflectivities of the gold and aluminum
mirrors were assumed to be flat in the wavelength range of interest (2-20 //m). The
178
6^
w G CO
ID N
'cS S o
:z,
1 0 1 1 1 2 1 3 A [yw.m]
m fl (C U H T) OJ N
O 2;
o 00
o CD
o
o N
-I -r r"! JMn
Z1 o 00
o CO
o
o CQ
,A, r-Ui,. i
1 1 1 1 . 5 1 2 1 2 . 5 1 5 1 6 1 7 1 8 1 9 2 0 A [/w.m] X \_l~iTci\
Figure D.l: Transmission profiles for the MIRAC L, N, 11.6, and Q., filters. Dashed
lines are the normalized filter transmission profiles. Solid lines are the relative
spectral response curves (= "RSRs", details given in Appendix A), and represent
the product of transmissions for the filters, detector, optics, KRS-5 dewar window,
and atmosphere. The RSRs were used for absolutely calibrating MIRAC on the
CWW system.
179
Table D.l. Zero-Magnitude Attributes of MIRAC Photometric Bands
(1) (2) (3) (4) (5) (6) (7) (8)
MIRAC Ajso Bandwidth In-Band FAOSO) Bandwidth Fj, (iso) i'(iso)
Band (urn) (fim) (Wcm-2) (W cm"" fim' •1) (Hz) ay) (Hz)
L 3.844 0.5423 2.631E-15 4.852E-15 1.102E+13 238.8 7.793E+13
uncert. 0.018 0.0037 1.609% 8.469E-17 6.820E+10 4.1 7.361E+11
N 10.35 3.228 3.263E-16 l.OllE-16 8.760E+12 37.25 2.946E+13
uncert. 0.05 0.022 1.632% 1.789E-18 4.170E+10 0.60 2.416E+11
11.6 11.57 0.8953 5.816E-17 6.496E-17 2.006E+12 29.00 2.592E+13
uncert. 0.08 0.0135 2.110% 1.686E-18 2.149E+10 0.61 2.827E+11
Qs 17.58 0.9130 1.123E-17 1.230E-17 8.834E+11 12.72 1.706E+13
uncert. 0.14 0.0185 2.494% 3.951E-19 1.263E+10 0.32 2.171E+11
Note. — Columns (1) name of MIRAC band, (2) isophotal wavelength, (3) wavelength
bandwidth of RSR, (4) In-band flux for zero-magnitude star, (5) isophotal monochro
matic intensity (wavelength units), (6) frequency bandwidth of RSR, (7) isophotal
monochromatic intensity (frequency imits), (8) isophotal frequency. Note that the stated
quantities assume that a KRS-5 dewar window is used. If the KBr dewar window is
used, the values are nearly identical. For KBr, every stated value is within 5% of the
stated uncertainty for the L, 11.6, and Qs bands, and within 36% of the stated imcer-
tainty for A' -band.
quantum efficiency for the MIRAC doped-silicon blocked-imp urity-band (BIB)
array was taken from Stapelbroek et al. (1995), following Hoffmann et al. (1998).
The zero magnitude attributes of the MIRAC filter systems are given in Table
D.l. Standard star fluxes on the CWW system for the four primary MIRAC filters
(with the KRS-5 dewar window) are given in Table D.2. When the KBr dewar
window is used on MIRAC-BLINC, the standard star fluxes are nearly identical
(to within <7% of the quoted flux uncertainties), so the same fluxes and magni
tudes can be safely adopted. The flux densities in Tables 3.4 and 3.5 are referenced
to this system.
180
Table D.2. Predicted MIRAC Standard Star Fluxes on CWW system
(1) (2) (3) (4) (5) (6) (7) (8) (9) (10)
HD Alt. Band Mag UIIC. FA unc. unc. unc.
Name Name (W cm~^ fj,m~ (Wcm~^/im~^) (mjy) (mJy) (%)
1522 t Cet L 0.800 0.022 2.32E-15 4.99E-17 1.14E+05 2.46E+03 2.15
1522 i Cet N 0.807 0.021 4.81E-17 9.98E-19 1.77E+04 3.68E+02 2.08
1522 t Cet 11.6 0.772 0.026 3.19E-17 9.01E-19 1.42E+04 4.02E+02 2.83
1522 L Cet Qs 0.775 0.030 6.02E-18 2.08E-19 6.23E+03 2.16E+02 3.46
12929 a Ari L -0.762 0.021 9.79E-15 2.01E-16 4.82E+05 9.88E+03 2.05
12929 Q Ari N -0.754 0.020 2.02E-16 4.00E-18 7.46E+04 1.47E+03 1.98
12929 a Ari 11.6 -0.789 0.025 1.34E-16 3.70E-18 6.00E+04 1.65E+03 2.75
12929 a Ari Qa -0.787 0.030 2.54E-17 8.63E-19 2.62E+04 8.93E+02 3.40
29139 a Tau L -3.045 0.021 8.01E-14 1.62E-15 3.94E+06 7.99E+04 2.03
29139 a Tau N -3.013 0.020 1.62E-15 3.21E-17 5.97E+05 1.18E+04 1,98
29139 a Tau 11.6 -3.074 0.025 l.lOE-15 3.03E-17 4.92E+05 1.35E+04 2.75
29139 a Tau <9. -3.058 0.029 2.06E-16 6.91E-18 2.13E+05 7.14E+03 3,36
45348 a Car L -1.289 0.019 1.59E-14 3.04E-16 7.83E+05 1.50E+04 1,91
45348 a Car N -1.309 0.020 3.38E-16 6.53E-18 1.24E+05 2.41E+03 1,94
45348 a Car 11.6 -1,307 0.025 2.17E-16 6.03E-18 9.67E+04 2.69E+03 2,78
45348 a Car Qs -1.307 0.029 4.10E-17 1.37E-18 4.24E+04 1.42E+03 3,35
48915 a CMa L -1.360 0.017 1.70E-14 2.96E-16 8.36E+05 1.46E+04 1,75
48915 a CMa N -1.348 0.018 3.50E-16 6.19E-18 1.29E+05 2.28E+03 1.77
48915 a CMa 11.6 -1.346 0.023 2.24E-16 5.82E-18 l.OOE+05 2.60E+03 2.59
48915 a CMa Qs -1.341 0.027 4.23E-17 1.36E-18 4.38E+04 1.41E+03 3.21
81797 a Hya h -1.362 0.019 1.70E-14 3.19E-16 8.37E+05 1.57E+04 1.88
81797 a Hya N -1.309 0.019 3.38E-16 6.39E-18 1.24E+05 2.36E+03 1.89
81797 a Hya 11.6 -1.351 0.027 2.25E-16 6.59E-18 l.OlE+05 2.94E+03 2.92
81797 a Hya Qs -1.350 0.030 4.26E-17 1.46E-18 4.41E+04 1.51E+03 3.42
106849 e Mus L -1.594 0.029 2.11E-14 5.83E-16 1.04E+06 2.87E+04 2.77
106849 £ Miis N -1.647 0.027 4.61E-16 1.17E-17 1.7GE+05 4.32E+03 2.54
106849 e Mtis 11.6 -1.708 0.031 3.13E-16 l.OlE-17 1.40E+05 4.53E+03 3.24
106849 € Mus Qs -1.700 0.039 5.89E-17 2.42E-18 6.09E-h04 2.50E+03 4.10
108903 7 Cru L -3.299 0.039 l.OlE-13 3.71E-15 4.99E+06 1.83E+05 3.66
108903 7 Cru N -3.354 0.038 2.22E-15 7.95E-17 8.18E+05 2.93E+04 3.58
Table D.2—Continued
(1) (2) (3) (4) (5) (6) (7) (8) (9) (10)
HD Alt. Band Mag imc. FA imc. Fu unc. unc.
Name Name (W cm" 2 (Wcm~^ (mjy) (mjy) (%)
108903 7 Cru 11.6 -3.413 0.041 1.51E-15 6.13E-17 6.73E+05 2.74E+04 4.07
108903 7 Cru Qs -3.403 0.043 2.83E-16 1.27E-17 2.92E+05 1.31E+04 4.49
128620 a Cen A L -1.562 0.018 2.04E-14 3.58E-16 l.OlE+06 1.76E+04 1.75
128620 a Cen A N -1.564 0.018 4.27E-16 7.57E-18 1.57E+05 2.79E+03 1.77
128620 a Cen A 11.6 -1.565 0.023 2.75E-16 7.13E-18 1.23E+05 3.18E+03 2.60
128620 a Cen A Q. -1.566 0.027 5.20E-17 1.67E-18 5.38E+04 1.73E+03 3.21
133216 a Lib L -1.565 0.025 2.05E-14 5.00E-16 l.OlE+06 2.46E+04 2.44
133216 a Lib N -1.619 0.022 4.49E-16 9.79E-18 1.66E+05 3.61E+03 2.18
133216 <T Lib 11.6 -1.680 0.028 3.05E-16 9.03E-18 1.36E+05 4.03E+03 2.96
133216 a Lib Q, -1.670 0.036 5.73E-17 2.23E-18 5.92E+04 2.30E+03 3.89
135742 /9Lib L 2.874 0.019 3.44E-16 6.39E-Z8 1.69E+04 3.14E+02 1.86
135742 /3 Lib N 2.899 0.019 7.00E-18 1.32E-19 2.58E+03 4.85E+01 1.88
135742 /3Lib 11.6 2.904 0.025 4.48E-18 1.23E-19 2.00E+03 5.49E+01 2.75
135742 ^Lib Qs 2.915 0.029 8.40E-19 2.78E-20 8.68E+02 2.88E+01 3.32
150798 a TrA L -1.337 0.022 1.66E-14 3.50E-16 8.18E+05 1.72E+04 2.11
150798 a TrA N -1.329 0.021 3.44E-16 6.99E-18 1.27E+05 2.58E+03 2.03
150798 a TrA 11.6 -1.364 0.026 2.28E-16 6.38E-18 1.02E+05 2.85E+03 2.80
150798 a TrA Qs -1.361 0.030 4.31E-17 1.48E-18 4.46E+04 1.53E+03 3.44
167618 TjSgr L -1.731 0.022 2.39E-14 5.08E-16 1.18E+06 2.50E+04 2.13
167618 rjSgz N -1.696 0.021 4.82E-16 9.86E-18 1.78E+05 3.63E+03 2.04
167618 j]Sgr 11.6 -1.753 0.026 3.26E-16 9.13E-18 1.46E+05 4.08E+03 2.80
167618 Tl Sgl Qs -1.786 0.031 6,37E-17 2.25E-18 6.59E+04 2.33E+03 3.53
216956 a PsA L 1.002 0.019 1.93E-15 3.62E-17 9.49E+04 1.78E+03 1.88
216956 a PsA N 1.001 0.019 4.02E-17 7.65E-19 1.48E+04 2.82E+02 1.90
216956 a PsA 11.6 1.004 0.025 2.58E-17 7.11E-19 1.15E+04 3.18E+02 2.76
216956 a PsA Qs 1.008 0.029 4.86E-18 1.62E-19 5.02E+03 1.67E+02 3.33
Note. — All stated quantities assume that a KRS-5 dewar window is used. If the KBr window is
used, the values are nearly identical (to within 7% of the stated uncertainties).
182
APPENDIX E
COMMENTS ON INDIVIDUAL SOURCES
The only young stars in our survey to show a significant iV-band excess were
the T Tauri stars HD 143006 and [PZ99] J161411.0-230536, both members of the
~5-Myr-old Upper Sco OB subgroup. Both stars are targets in the FEPS Spitzer
Legacy Science program, but only HD 143006 was previously known to possess
a circumstellar disk. Both were detected by IRAS, and the MIRAC jV-band fluxes
are consistent with the color-corrected llfim measurements in the IRAS FSC
(Moshir et al., 1990). Here we discuss these stars in more detail.
E.l [PZ99] J161411.0-230536
J161411 is a KO-type weak-lined T Tauri (EW(Ha) = 0.96A) star discovered by
Preibisch et al. (1998). In a spectroscopic survey to identify new members of Up
per Sco (Mamajek, Meyer, & Liebert, in prep.), the authors obtained a red, low-
resolution spectrum of J161411 in July 2000, which shows an asymmetric Ha fea
ture with blueshifted emission and redshifted absorption (net EW(Ha) = 0.36A).
We confirm the strong lithium absorption (EW(Li A6707) = 0.45 A) observed by
Preibisch et al. (1998). The UCAC2 proper motion (Zacharias et al., 2003) for
J161411 is consistent with membership in the Upper Sco subgroup. The {KG — N)
color (=2.1) is similar to that of classical T Tauri stars in Taurus-Auriga (Kenyon
& Hartmann, 1995). The photometric and spectroscopic evidence suggest that
this ~5-Myr-old, ~1 M© star is actively accreting from a circumstellar disk.
183
E.2 HD 143006
HD 143006 is a G5Ve (Henize, 1976) T Tauri star with strong Li absorption (EW(Li
A6707) = 0.24A; Dunkin, Barlow, & Ryan, 1997). The star is situated in the middle
of the Upper Sco OB association, and its proper motion (na, ns = -H, -20 mas/yr;
Zacharias et al., 2003) and radial velocity (-0.9 km/s; Dunkin, Barlow, & Ryan,
1997) are indistinguishable from other association members (de Bruijne, 1999b).
Several studies have classified HD 143006 as a distant G-type supergiant or "pre-
planetary nebula" (Carballo, Wesselius, & Whittet, 1992; Kohoutek, 2001), how
ever we believe this is erroneous. If HD 143006 were indeed a supergiant at d =
3.4 kpc (Pottasch & Parthasarathy, 1988), its tangential velocity would be ~370
km/s - extraordinarily fast for a population I star. The MIRAC N and 11.6 pho
tometry agrees well with the data points in the SED for HD 143006 plotted in Fig.
2 of Sylvester et al. (1996). The spectral energy distribution for HD 143006 and its
optically-thick disk is well-studied from 0.4-1300 //m, so we do not discuss this
object further.
184
REFERENCES
Adams, F. C., Lada, C. J., & Shu, F. H. 1987, ApJ, 312,788
Agnor, C. B., Canup, R. M., & Levison, H. F. 1999, Icarus, 142,219
Alcala, J. M., Krautter, J., Schmitt, J. H. M. M., Co vino, E., Wichmann, R., &
Mundt, R. 1995, A&AS, 114,109
Arias, E. F., Chariot, R, Feissel, M., & Lestrade, J.-F. 1995, A&A, 303,604
Artymowicz, R 1988, ApJ, 335, L79
Asiain, R., Figueras, E, Torra, J., & Chen, B. 1999, A«&;A, 341,427
Atanasijevic, I. 1971, Selected Exercises in Galactic Astronomy, Astrophysics and
Space Science Library, Dordrecht: Reidel, 1971
Aumann, H. H. & Probst, R. G. 1991, ApJ, 368,264
Backman, D. E. & Paresce, F. 1993, Protostars and Planets III, 1253
Ballesteros-Paredes, J., Hartmann, L., & V^zquez-Semadeni, E. 1999, ApJ, 527,
285
Balona, L. A., 1994, MNRAS, 268,119
Basri, G., Marcy, G. W., & Graham, J. R., 1996, ApJ, 458,600
Beck with, S. V. W. 1999, NATO ASIC Proc. 540: The Origin of Stars and Planetary
Systems, 579
Beckwith, S. V. W., Sargent, A. L, Chini, R. S., & Guesten, R. 1990, AJ, 99,924
185
Beichman, C. A., Myers, P. C., Emerson, J. P., Harris, S., Mathieu, R., Benson, P. J.,
& Jennings, R. E. 1986, ApJ, 307,337
Beichman, C. A., Neugebauer, G., Habing, H. J., Clegg, P. E., & Chester, T. J. 1988,
NASA RP-1190, Vol. 1
Bertelli, G., Bressan, A., Chiosi, C., Fagotto, F., & Nasi, E., 1994, A&AS, 106,275
Bertout, C., 1989, ARA&A, 27,351
Bertout, C., Robichon, N., & Arenou, F. 1999, A&A, 352,574
Bessell, M.S., 2000, PASP, 112,961
Bessell, M. S. & Brett, J. M. 1988, PASP, 100,1134
Bessell, M.S., Castelli, F., & Plez, B., 1998, A&A, 333,231
Bhatt, H. C., 2000, A&A, 362,715
Binney, J. & Merrifield, M. 1998, Galactic Astronomy, Princeton, NJ : Princeton
University Press.
Blaauw, A., 1946, Ph.D. Thesis, Gronningen University
Blaauw, A., 1991, in The Physics of Star Formation and Early Stellar Evolu
tion, NATO ASl Series C, Vol. 342, eds. C.J. Lada & N.D. Kylafis (Dordrecht:
Kluwer), 125
Blitz, L. & Shu, F. H. 1980, ApJ, 238, 148
Boesgaard, A. M., Budge, K. G., & Ramsay, M. E. 1988, ApJ, 327,389
Bouvier, J., Wichmann, R., Grankin, K., Allain, S., Covino, E., Fernandez, M.,
Martin, E. L., Terranegra, L., Catalano, S., Marilli, E., 1997, A&A, 318,495
186
Briceno, C., Hartmann, L. W., Stauffer, J. R., Gagne, M., Stem, R. A., & Caillault,
J. 1997, AJ, 113,740
Briggs, R. E. 1962, AJ, 67, 710
Bums, J. A., Lamy, P. L., & Soter, S. 1979, Icams, 40,1
Butler, R. P., Marcy, G. W., Cohen, R. D., & Duncan, D. K. 1987, ApJL, 319, L19
Carballo, R., Wesselius, P. R., & Whittet, D. C. B. 1992, A&A, 262,106
Carpenter, J. M. 2001, AJ, 121,2851
Cayrel de Strobel, G. & Cayrel, R. 1989, A&A, 218, L9
Cayrel de Strobel, G., Soubiran, C., Ralite, N., 2001, A&A, 373,159
Chambers, J. E. 2001, Icarus, 152,205
Chen, B., Asiain, R., Figueras, E, & Torra, J. 1997, A&A, 318,29
Chen, C. H. & Jura, M. 2001, ApJ, 560, LI 71
Chereul, E., Creze, M., & Bienayme, 0.1999, A&AS, 135,5
Chiang, E. L & Goldreich, P 1997, ApJ, 490,368
Clark, R O. 1996, PLEXUS Version 2.1a, CD-ROM (Hanscom AFB, MA; Phillips
Lab., Dir. Geophys., Air Force Mater. Command).
Cohen, M. & Kuhi, L. V. 1979, ApJS, 41,743
Cohen, M., Schwartz, D. E., Chokshi, A., & Walker, R. G. 1987, AJ, 93,1199
Cohen, M., Walker, R. G., Barlow, M. J., & Deacon, J. R. 1992, AJ, 104,1650
187
Cohen, M., Walker, R. G., Carter, B., Hammersley, P., Kidger, M., & Noguchi, K.
1999, AJ, 117,1864
Cohen, M, Megeath, S. T., Hammersley, P. L., Martin-Luis, R, & Stauffer, J. 2003,
AJ, 125,2645
Cohen, M., Wheaton, W. A., & Megeath, S. T. 2003, AJ, 126,10
Covino, E., Alcala, J. M., Allain, S., Bouvier, J., Terranegra, L., & Krautter, J., 1997,
A&A, 328,187
Crawford, D. L., 1978, AJ, 83,48
Crawford, D. L. & Mandwewala, N., 1976, PASP, 88,917
Cutispoto, G., Pastori, L., Pasquini, L., de Medeiros, J. R., Tagliaferri, G., & An
dersen, J. 2002, A&A, 384,491
Cutri, R. M. et al. 2003, 2MASS All-Sky Catalog of Point Sources, VizieR Online
Data Catalog, 2246,0
D'Antona, F. & Mazzitelli, L, 1997, Mem. Soc. Astr. Ital., 68, 807 (DM97)
de Bruijne, J. H. J., 1999a, MNRAS, 306,381
de Bruijne, J. H. J., 1999b, MNRAS, 310,585
de Geus, E., 1992, A&A, 262,259
de Geus, E., de Zeeuw, P. & Lub, J., 1989, A&A, 216,44
Dehnen, W. 1998, AJ, 115,2384
de Jager, C. & Nieuwenhuijzen, H., 1987, A&A, 177,217
188
de Zeeuw, P. T. & Brand, J., 1985, in Birth and Evolution of Massive Stars and Stel
lar Groups; Proceedings of the Symposium, Dwingeloo, Netherlands, Septem
ber 24-26,1984, (Dordrecht: D. Reidel), 102
de Zeeuw, P. T., Hoogerwerf, R., de Bruijne, J. H. J., Brown, A. G. A., & Blaauw,
A., 1999, AJ, 117,354 (dZ99)
Dohnanyi, J. W. 1969, J. Geophys. Res., 74,2531
Dominik, C. & Decin, G. 2003, ApJ, 598, 626
Drilling, J. S., & Landolt, A. U., 2000, in Allen's Astrophysical Quantities, 4th. Ed.,
ed. Arthur C. Cox, (New York: AIP Press, Springer)
Dunkin, S. K., Barlow, M. J., & Ryan, S. G. 1997, MNRAS, 290,165
Eggen, O. J. 1961, Royal Greenwich Observatory Bulletin, 41,245
Eggen, O. J. 1975, PASP, 87,37
Eggen, O. J. 1995, AJ, 110,1749
Eggen, O. J. 1998, unpublished manuscript
ESA 1997, The Hipparcos and Tycho Catalogues, ESA SP-1200
Evans, N. J. 1999, ARA&A, 37,311
Fajardo-Acosta, S. B., Beichman, C. A., & Cutri, R. M. 2000, ApJ, 538, LI55
Feast, M. & Whitelock, P. 1997, MNRAS, 291,683
Feigelson, E. D., 1996, ApJ, 468,306
Eeigelson, E. D. & Kriss, G. A. 1981, ApJL, 248, L35
189
Fixsen, D. J. & Dwek, E. 2002, ApJ, 578,1009
Fleming, T. A., Molendi, S., Maccacaro, T., & Wolter, A., 1995, ApJS, 99,701
Frink, S. 1999, Ph.D. Thesis, Astronomisches Rechen-Institut Heidelberg
Frogel, J. A., & Stothers, R., 1977, AJ, 82,890
Gaidos, E. J. 1999, ApJ, 510, L131
Garcia, B., 1989, Bull. Inform. CDS 36,27
Garcia Lopez, R. J., Rebolo, R., & Martin, E. L. 1994, A&A, 282,518
Gehrels, N., 1986, ApJ, 303,336
Ghez, A. M., Neugebauer, G., & Matthews, K. 1993, AJ, 106,2005
Golay, M. 1974, Astrophysics and Space Science Library, 41
Gontcharov, G. A., Andronova, A. A., Titov, O. A., & Kornilov, E. V., 2001, A&A,
265,222
Gray, D. E, 1991, The Observation and Analysis of Stellar Photospheres, 2nd ed.,
(New York: Cambridge)
Gray, R.O., 2000, A Digital Spectral Classification Atlas,
http: / / nedwww.ipac.caltech.edu/level5/Gray/
Gray, R.O., Graham, RW., & Hoyt, S.R., 2001, AJ, 121,2159
Greenberg, R. & Nolan, M. C. 1989, Asteroids II, 778
Gregorio-Hetem, J., Lepine, J. R. D., Quast, G. R., Torres, C. A. O., & de la Reza,
R., 1992, AJ, 103, 549
190
Griffin, R. F. 1998, The Observatory, 118,223
Grocholski, A. J. & Sarajedini, A. 2003, MNRAS, 345,1015
Griin, E., Zook, H. A., Fechtig, H., & Giese, R. H. 1985, Icarus, 62,244
Guillout, P., Sterzik, M. F., Schmitt, J. H. M. M., Motch, C., & Neuhauser, R., 1998,
A&A, 337,113
Guillout, P., Schmitt, J. H. M. M., Egret, D., Voges, W., Motch, C., & Sterzik, M. F.
1999, A&A, 351,1003
Habing, H. J. et al. 2001, A&A, 365,545
Haisch, K. E., Lada, E. A., & Lada, C. J. 2001a, ApJ, 553, L153
Haisch, K. E., Lada, E. A., & Lada, C. J. 2001b, AJ, 121,2065
Hamuy, M., Suntzeff, N. B., Heathcote, S. R., Walker, A. R., Gigoux, P., & Phillips,
M. M., 1994, PASP, 106,566
Harper, D. A., Loewenstein, R. E, & Davidson, J. A. 1984, ApJ, 285,808
Hartigan, P., Edwards, S., & Ghandour, L., 1995, ApJ, 452,736
Hartmann, L. 2001, AJ, 121,1030
Hartmann, L., Ballesteros-Paredes, J., & Bergin, E. A. 2001, ApJ, 562,852
Hartmann, W. K. & Davis, D. R. 1975, Icarus, 24,504
Hartmann, L., Stauffer, J. R., Kenyon, S. J., & Jones, B. F. 1991, AJ, 101,1050
Hauck, B. & Mermilliod, M., 1997, A&AS, 129,431
Hawley, S. L., Gizis, J. E., & Reid, L N. 1996, AJ, 112,2799
191
Henize, K. G. 1976, ApJS, 30,491
Herbig, G. H., 1978, in Problems of Physics and Evolution of the Universe, ed. L.
V. Mirzoyan (Yervan: Acad. Sci. Armenian SSR), 171
Herbig, G. H., Bell, K. Robin, 1988, Lick Obs. Bulletin No. 1111, (Santa Cruz: Lick
Observatory)
Herbig, G. H., Vrba, F. J., & Rydgren, A. E. 1986, AJ, 91,575
Hillenbrand, L. A., Strom, S. E., Calvet, N., Merrill, K. M., Gatley, I., Makidon,
R. B., Meyer, M. R., & Skmtskie, M. F. 1998, AJ, 116,1816
Hillenbrand, L. A., et al. 2004, in prep
Hillenbrand, L. A. & Meyer, M. R., 1999, BAAS, 195, #209
Hinz, R M., Angel, J. R. R, Woolf, N. J., Hoffmann, W. R, & McCarthy, D. W. 2000,
Proc. SPIE, 4006,349
Hoffmann, W. F., Hora, J. L., Fazio, G. G., Deutsch, L. K., & Dayal, A. 1998,
Proc. SPIE, 3354,647
Hoffmann, W. F. & Hora, J. L. 1999, MIRAC3 User's Manual, http;//cfa-
www.harvard.edu/~jhora/mirac/ mrcman.pdf
H0g, E., Kuzmin, A., Bastian, U., Fabricius, C., Kuimov, K., Lindegren, L.,
Makarov, V.V., & Roeser, S., 1998, A&A, 335, L65
H0g, E., Fabricius, C., Makarov, V.V., Urban, S., Corbin, T., Wycoff G., Bastian U.,
Schwekendiek P., Wicenec A. 2000a, A&A, 355, L27
H0g, E., Fabricius, C., Makarov, V.V., Bastian, U., Schwekendiek, P., Wicenec, A.,
Urban, S., Corbin, T., Wycoff, G. 2000b, A&A, 357,367
192
Hollenbach, D. J., Yorke, H. W., & Johnstone, D. 2000, Protostars and Planets IV,
401
Hoogerwerf, R. 2000, MNRAS, 313,43
Hoogerwerf, R. & Aguilar, L. A., 1999, MNRAS, 306,394
Hoogerwerf, R., de Bruijne, J. H., J., & de Zeeuw, P. T., 2001,365,49
Hora, J. L. 1991, Ph.D. Thesis, The University of Arizona
Houk, N., 1978, Michigan Catalogue of Two-Dimensional Spectral Types for HD
Stars. (Vol. 2, -52° < -40°), (Ann Arbor: Univ. Michigan)
Houk, N., 1982, Michigan Catalogue of Two-Dimensional Spectral Types for HD
Stars. (Vol. 3, -40° < -26°), (Ann Arbor: Univ. Michigan)
Houk, N. & Cowley, A. P., 1975, Michigan Catalogue of Two-Dimensional Spec
tral Types for HD Stars. (Vol. 1, -90° < -53°), (Ann Arbor: Univ. Michigan)
Houk, N. & Smith-Moore, M., 1988, Michigan Catalogue of Two-Dimensional
Spectral Types for HD Stars. (Vol. 4, -26° < -12°), (Ann Arbor: Univ. Michigan)
Houk, N. & Swift, C., 1999, Michigan Catalog of Two-Dimensional Spectral Types
for the HD Stars. (Vol. 5., -12° < +5°), (Ann Arbor: Univ. Michigan)
Hughes, J., Hartigan, P., Krautter, J., & Kelemen, J., 1994, AJ, 108,1071
Hughes, J., Hartigan, P., Clampitt, L., 1993, AJ, 105, 571
IPAC, 1986, IRAS Point Source Catalog
Jayawardhana, R., Hartmann, L., Fazio, G., Fisher, R. S., Telesco, C. M., & Pina,
R. K. 1999, ApJ, 521, L129
193
Jeffries, R. D. 1995, MNRAS, 273,559
Jeffries, R. D. 1999, MNRAS, 309,189
Jeffries, R. D. & Jewell, S. J. 1993, MNRAS, 264,106
Jensen, E., 2001, in Young Stars Near Earth: Progress & Prospects, eds. R.
Jayawardhana & T. Greene, ASP Conf. Ser. Vol. 244, (San Francisco: ASP), 3
Johnson, H. L, & Morgan, W. W., 1953, ApJ, 117,313
Jones, B. F., Shetrone, M., Fischer, D., & Soderblom, D. R. 1996, AJ, 112,186
Jones, B. F., Fischer, D., Shetrone, M., & Soderblom, D. R., 1997, AJ, 114,352
Jones, B. F. & Herbig, G. H. 1979, AJ, 84,1872
Jourdain de Muizon, M., et al. 1999, A&A, 350,875
Joy, A. H. 1945, ApJ, 102,168
Jura, M., Zuckerman, B., Becklin, E. E., & Smith, R. C. 1993, ApJL, 418, L37
Kastner, J. H., Zuckerman, B., Weintraub, D. A., &Forveille, T., 1997, Science, 227,
67
Keenan, PC., 1984, in The MK Process and Stellar Classification, ed. R. F. Garri
son, (Toronto: Univ. of Toronto), 29
Keenan, PC. & Bambaum, C., 1999, ApJ, 518,859
Keenan, PC. & McNeil, R.C. 1976, An Atlas of Spectra of the Cooler Stars,
(Columbus: Ohio State Univ.)
Keenan, PC. & McNeil, R.C., 1989, ApJS, 71,245
194
Keenan, P.C. & Yorka, S.B., 1988, Bull. Inform. CDS 35,37
Kelsall, T. et al. 1998, ApJ, 508,44
Kenyon, S. J. & Bromley, B. C. 2004, ApJ, 602, LI33
Kenyon, S. J. & Hartmann, L. 1995, ApJS, 101,117
Kleine, X, Miinker, C., Mezger, K., & Palme, H. 2002, Nature, 418,952
Kleine, T., Mezger, K., & Miinker, C. 2003, Meteoritics & Planetary Science, vol. 38,
Supplement, abstract no.5212,38, 5212
Kleinmarm, S. G., Cutri, R. M., Young, E. T., Low, F. J., & Gillett, F. C. 1986, IRAS
Serendipitous Survey Catalog, Tucson, Univ. of Arizona
Koenigl, A. 1991, ApJL, 370, L39
Kohler, R., Kunkel, M., Leinert, C., & Zinnecker, H., 2000, A&A, 356,541
Kohler, R. 2001, AJ, 122,3325
Kohoutek, L. 2001, A&A, 378,843
Krautter, J., Wichmann, R., Schmitt, J.H.M.M., Alcala, J.M., Neuhauser, R., & Ter-
ranegra, L., 1997, A&AS, 123,329
Kroupa, R, 2001, MNRAS, 322,231
Lachaume, R., Doniinik, C., Lanz, T., & Habing, H. J. 1999, A&A, 348, 897
Lada, C. J. 1987, lAU Symp. 115: Star Forming Regions, 115,1
Lada, C. J. & Wilking, B. A. 1984, ApJ, 287,610
195
Lagrange, A.-M., Backman, D. E., & Artymowicz, P. 2000, Protostars and Planets
IV, 639
Lawson, W. A., Crause, L. A., Mamajek, E. E., & Feigelson, E. D. 2002, MNRAS,
329, L29
Lawson, W. A., Lyo, A.-R., & Muzerolle, J., 2004, MNRAS, in press
Lesh, J. R. 1968, ApJS, 17,371
Lindblad, P. O., Palous, J., Loden, K., & Lindegren, L., 1997, ESA SP-402: Hippar-
cos - Venice '97,402,507
Lindegren, L., Madsen, S., & Dravins, D. 2000, A&A, 356,1119
Lowrance, P. J., Schneider, G., Kirkpatrick, J. D., Becklin, E. E., Weinberger, A. J.,
Zuckerman, B., Plait, P., Malmuth, E. M., Heap, S. R., Schultz, A., Smith, B. A.,
Terrile, R. J., & Hines, D. C., 2000, ApJ, 541, 390
Lyo, A.-R., Lawson, W. A., Mamajek, E. E., Feigelson, E. D., Sung, E., & Crause,
L. A. 2003, MNRAS, 338,616
Madsen, S., Dravins, D., & Lindegren, L. 2002, A&A, 381,446
Maeder, A., 1981, A&A, 101,385
Makarov, V. V. 2003, AJ, 126,1996
Malyuto, V. & Schmidt-Kaler, Th. 1997, A&A, 693, 699
Mamajek, E.E., 2003, proceedings of "Open Issues in Local Star Formation and
Early Stellar Evolution", Ouro Preto (astro-ph/0305209)
196
Mamajek, E.E., Meyer, M.R., Hinz, P.M., Hoffmann, W.F., Cohen, M., Hora, J.L.,
2004, ApJ, in press (astro-ph/0405271)
Mamajek, E.E., Feigelson, E.D., 2001, Young Stars Near Earth: Progress &
Prospects, ASP Conf. Ser. Vol. 244, eds. R. Jayawardhana & T. Greene, (San
Francisco: ASP), 104.
Mamajek, E.E., Lawson, W.A. & Feigelson, E.D. 1999, ApJ, 516, L77
Mamajek, E. E., Meyer, M. R., & Liebert, J. 2002, AJ, 124,1670
Marcy, G. W. & Butler, R. P 2000, PASP, 112,137
Marsakov, V. A. & Shevelev, Y. G. 1995, Bulletin d'Information du Centre de Don-
nees Stellaires, 47,13
Mashonkina, L. & Gehren, T., 2001, A&A, 376, 232
Mathioudakis, M. & Doyle, J. G. 1993, A&A, 280,181
Mathis, J. S., 1990, ARA&A, 28,37
Ma this, J. S., Rumpl, W., & Nordsieck, K. H. 1977, ApJ, 217,425
McCaughrean, M. J. & O'Dell, C. R. 1996, AJ, 111, 1977
McKee, C. F., Zweibel, E. G., Goodman, A. A., & Heiles, C. 1993, Protostars and
Planets III, 327
Melo, C. H. F. 2003, A&A, 410,269
Mendoza, E. E. 1968, ApJ, 151,977
Metchev, S. A., Hillenbrand, L. A., & Meyer, M. R. 2004, ApJ, 600,435
197
Meyer, M. R., Calvet, N., & Hillenbrand, L. A., 1997, AJ, 114,288
Meyer, M. R. et al. 2002, The Origins of Stars and Planets: The VLT View. Pro
ceedings of the ESO Workshop held in Garching, Germany, 24-27 April 2001,
p. 463.
Meyer, M. R. et al., 2004, ApJS, in press
Montmerle, T., Koch-Miramond, L., Falgarone, E., & Grindlay, J. E. 1983, ApJ,
269,182
Moreno, E., Alfaro, E. J., & Franco, J., 1999, ApJ, 522,276
Moshir, M. & et al. 1990, IRAS Faint Source Catalogue, version 2.0
Motte, F., Andre, P., & Neri, R. 1998, A&A, 336,150
Mukai, T. & Yamamoto, T. 1982, A&A, 107,97
Mundt, R., Walter, F. M., Feigelson, E. D., Finkenzeller, U., Herbig, G. H., & Odell,
A. R 1983, ApJ, 269,229
Muzerolle, J., Calvet, N., Briceno, C., Hartmann, L., & Hillenbrand, L. 2000, ApJ,
535, L47
Muzerolle, J., Calvet, N., Hartmann, L., & D'Alessio, P. 2003, ApJ, 597, L149
Muzerolle, J., Hartmann, L., & Calvet, N., 1998, AJ, 116,455
Myers, P. C. 1999, NATO ASIC Proc. 540: The Origin of Stars and Planetary Sys
tems, 67
Napiwotzki, R., Schonberner, D., & Wenske, V., 1993, A&A, 268,653
198
Nelson, G. J., Robinson, R. D., Slee, O. B., Ashley, M. C. B., Hyland, A. R., Tuohy,
I. R., Nikoloff, I., & Vaughan, A. E. 1986, MNRAS, 220,91
Nordstrom, B., et al. 2004, A&A, 418,989
Palla, F. & Galli, D. 1997, ApJL, 476, L35
Palla, R, & Stabler, S. W., 2001, ApJ, 553,299 (PSOl)
Pallavicini, R., Cerruti-Sola, M., & Duncan, D.K., 1987, A&A, 174,116
Patten, B. H., & Simon, T., 1996, ApJS, 106,489
Paulson, D. B., Sneden, C., & Cochran, W. D. 2003, AJ, 125,3185
Perryman, M. A. C. 2003, ASP Conf. Ser. 298: GAIA Spectroscopy: Science and
Technology, 3
Pilachowski, C. A., Booth, J., & Hobbs, L. M. 1987, PASP, 99,1288
Pinsonneault, M. H., Terndrup, D. M., Hanson, R. B., & Stauffer, J. R. 2004, ApJ,
600,946
Pojmanski, G., 1998, Acta Astron., 48,35
Poppel, W., 1997, Fund. Cosmic Phys., 18,1
Popper, D. M., 1966, AJ, 71,175
Pottasch, S. R. & Parthasarathy, M. 1988, A&A, 192,182
Preibisch, T., Guenther, E., Zinnecker, H., Sterzik, M., Frink, S., & Roeser, S. 1998,
A&A, 333,619
199
Preibisch, T., Brown, A. G. A., Bridges, T., Guenther, E., & Zinnecker, H. 2002, AJ,
124,404
Preibisch, T. & Zinnecker, H., 1999, AJ, 117,2381
Preibisch, T. & Zinnecker, H., 2000, in Stellar Clusters and Associations: Con
vection, Rotation, and Dynamos, ASP Conf., Vol.198., eds. by R. Pallavicini, G.
Micela, and S. Sciortino, 219
Press, et al. 1986, Numerical Recipes in Fortran 77,2nd Edition, Vol. 1 of Fortran
Numerical Recipes, Cambridge University Press, New York
Pye, J. P, et al. 1995, MNRAS, 274,1165
Probst, R. G. 1983, ApJS, 53,335
Randich, S., Schmitt, J. H. M. M., Prosser, C. F., & Stauffer, J. R., 1995, A&A, 300,
134
Randich, S., Aharpour, N., Pallavicini, R., Prosser, C. R, & Stauffer, J. R., 1997,
A&A, 323,86
Randich, S., Gratton, R., Pallavicini, R., Pasquini, L., & Carretta, E., 1999, A&A,
348,487
Randich, S., Pallavicini, R., Meola, G., Stauffer, J. R., & Balachandran, S. C. 2001,
A&A, 372,862
Reach, W. T., Morris, P., Boulanger, F., & Okumura, K. 2003, Icarus, 164,384
Rebull, L. M., Wolff, S. C., Strom, S. E., & Makidon, R. B., 2001, ApJ, in press
Reid, I. N., Kilkenny, D., & Cruz, K. L. 2002, AJ, 123,2822
200
Rodgers, A. W., Conroy, P., & Bloxham, G., 1988, PASP, 100,626
Rose, J.A., 1984, AJ, 89,1238
Sarajedini, A., Brandt, K., Grocholski, A. J., & Tiede, G. P. 2004, AJ, 127,991
Sartori, M. J., Lepine, J. R. D., & Dias, W. S. 2003, A&A, 404,913
Schmidt-Kaler, Th., 1982, in Landolt-Bomstein: Numerical Data and Functional
Relationships in Science and Technology, eds. K. Schaifers & H. H. Voigt,
(Berlin; Springer-Verlag)
Shobbrook, R. R., 1983, MNRAS, 205,1215
Shu, F. H., Adams, F. C., & Lizano, S. 1987, ARA&A, 25,23
Shu, F. H., Allen, A., Shang, H., Ostriker, E. C., & Li, Z. 1999, NATO ASIC
Proc. 540: The Origin of Stars and Planetary Systems, 193
Siess, L., Dufour, E. & Forestini, M., 2000, A&A, 358,593 (SDFOO)
Siess, L., Forestini, M., & Dougados, C. 1997, A&A, 324,556
Silverstone, M. D. 2000, Ph.D. Thesis, UCLA
Skrutskie, M. F., Dutkevitch, D., Strom, S. E., Edwards, S., Strom, K. M., & Shure,
M. A. 1990, AJ, 99,1187
Skuljan, J., Hearnshaw, J. B., & Cottrell, P. L. 1999, MNRAS, 308, 731
Slawson, R. W., Hill, R. J., & Landstreet, J. D., 1992, ApJS, 82,117
Smart, W. M., 1968, Stellar Kinematics, (London:Longmans)
201
Soderblom, D. R., Jones, B. R, Balachandran, S., Stauffer, J. R., Duncan, D. K.,
Fedele, S. B., & Hudon, J. D. 1993, AJ, 106,1059
Soderblom, D. R., King, J. R., & Henry, T. J., 1998, AJ, 116,396
Song, I., Weinberger, A. J., Becklin, E. E., Zuckerman, B., & Chen, C. 2002, AJ, 124,
514
Song, L, Zuckerman, B., & Bessell, M. S. 2003, ApJ, 599,342
Spangler, C., Sargent, A. I., Silverstone, M. D., Becklin, E. E., & Zuckerman, B.
2001, ApJ, 555,932
Spergel, D. N., et al. 2003, ApJS, 148,175
Stapelbroek, M. G., Seib, D. H., Huffman, J. E., & Florence, R. A. 1995, Proc. SPIE,
2475,41
Stauffer, J. R., Caillault, J.-R, Gagne, M., Prosser, C. P., & Hartmann, L. W., 1994,
ApJS, 91, 625
Stauffer, J. R., Schultz, G., & Kirkpatrick, J. D. 1998, ApJL, 499, LI99
Stelzer, B. & Neuhauser, R. 2001, A&A, 372,117
Sterzik, M. P., & Durisen, R. H., 1995, A&A, 304, L9
Stevenson, D. J. 1987, Annual Review of Earth and Planetary Sciences, 15,271
Strom, K. M., Wilkin, P. P., Strom, S. E., & Seaman, R. L. 1989, AJ, 98,1444
Sung, H., Bessell, M. S., Lee, B., & Lee, S. 2002, AJ, 123,290
Sylvester, R. J., Skinner, C. J., Barlow, M. J., & Mannings, V. 1996, MNRAS, 279,
915
202
Tachihara, K., Toyoda, S., Onishi, T., Mizuno, A., Fukui, Y., & Neuhauser, R., 2002,
PASJ, in press.
Taylor, B. J. 2000, A&A, 362,563
Torra, J., Fernandez, D., & Figueras, F., 2000, A&A, 359,82
Torres, C. A. O., da Silva, L., Quast, G. R., de la Reza, R., & Jilinski, E. 2000, AJ,
120,1410 (TDQOO)
Torres, C. A. O., da Quast, G. R., de la Reza, R., da Silva, L., & Melo, C. H. F. 2001,
ASP Conf. Ser. 244; Young Stars Near Earth, 43.
Torres, C. A. O., Quast, G. R., de La Reza, R., da Silva, L., & Melo, C. H. F. 2003,
ASP Conf. Ser. 287: Galactic Star Formation Across the Stellar Mass Spectrum,
439
Trumpler, R. J. & Weaver, H. E, 1953, Statistical Astronomy, Dover Books on As
tronomy and Space Topics, New York: Dover Publications
Urban, S.E., Corbin, T.E., & Wycoff, G.L. 1998, AJ, 115,2161
van der Bliek, N. S., Manfroid, J., & Bouchet, P. 1996, A&AS, 119,547
Voges, W., et al., 1999, A&A, 349,389
Wallace, L., Hinkle, K., & Livingston, W. 1998, An Atlas of the Spectrum of the
Solar Photosphere from 13,500 to 28,000 cm""' (3570 to 7405 A),(Tucson; NOAO)
Walter, F. M., Brown, A., Mathieu, R. D., Myers, P. C., & Vrba, F. J. 1988, AJ, 96,
297
Walter, E M. & Kuhi, L. V. 1981, ApJ, 250,254
203
Waters, L. B. F. M., Cote, J., & Aumann, H. H. 1987, A&A, 172,225
Webb, R., A., Zuckerman, B., Platais, I., Patience, ]., White, R. J., Schwartz, M. J.,
McCarthy, C., 1999, ApJ, 512, L63
Weidenschilling, S. J. 1977, Ap&SS, 51,153
Weinberger, A. J., Becklin, E. E., Zuckerman, B., & Song, 1.2003, BAAS, 202, #34.01
Weinberger, A. J., Becklin, E. E., Zuckerman, B., & Song, 1.2003, BAAS, 203, #13.03
Westin, T. N. G., 1985, A&AS, 60,99
Wetherill, G. W. & Stewart, G. R. 1993, Icarus, 106,190
Whittet, D. C. B. 2003, Dust in the Galactic Environment, 2nd ed., Bristol: Institute
of Physics (lOP) Publishing, Series in Astronomy and Astrophysics
Wichmann, R., et al. 1996, A&A, 312,439
Wichmann, R., Krautter, J., Covino, E., Alcala, J. M., Neuhauser, R., Schmitt, J. H.
M. M., 1997, A&A, 320,185
Wichmann, R., Sterzik, M., Krautter, J., Metanomski, A., & Voges, W., 1997b,
A&A, 326,211
Wichmann, R., Bastian, U., Krautter, J., Jankovics, I., & Rucinski, S. M., 1998,
MNRAS, 301, L39
Wichmann, R., et al. 2000, A&A, 359,181
Wichmann, R. & Schmitt, J. H. M. M. 2003, MNRAS, 342,1021
Wichmann, R., Schmitt, J. H. M. M., Hubrig, S., 2003, A&A, 399,983
204
Wilking, B. A., Lada, C. J., & Young, E. T., 1989, ApJ, 340,823
Wolfe, W. L. & Zissis, G. J. 1985, Arlington: Office of Naval Research, Department
of the Navy
Wolf, S. & Hillenbrand, L. A. 2003, ApJ, 596,603
Wolk, S. J. & Walter, F. M. 1996, AJ, 111, 2066
Wood, B. E., Miiller, H., Zank, G. R, & Linsky, J. L. 2002, ApJ, 574,412
Zacharias, N., Urban, S. E., Zacharias, M. 1., Wycoff, G. L., Hall, D. M., Germain,
M. E., Holdenried, E. R., & Winter, L. 2003, VizieR Online Data Catalog, 1289,0
Zuckerman, B., Song, I., Bessell, M. S., & Webb, R. A. 2001, ApJ, 562, L87
Zuckerman, B., Song, I. 2004, ApJ, 603, 738
Zuckerman, B., Song, I., Bessell, M. S., & Webb, R. A. 2001, ApJ, 562, L87
Zuckerman, B., Song, L, & Webb, R. A., 2001, ApJ, 559,388 (ZSWOl)
Zuckerman, B. & Webb, R. A., 2000, ApJ, 535,959 (ZWOO)
top related